978 1 4471 2867 0

Designing Inclusive Systems
Patrick Langdon John Clarkson

•
Peter Robinson Jonathan Lazar

•
Ann Heylighen
Editors
Designing Inclusive Systems

Designing Inclusion for Real-world
Applications
123
Patrick Langdon Jonathan Lazar
Department of Engineering Universal Usability Laboratory
Cambridge Engineering Design Centre Department of Computer and Information
University of Cambridge Sciences
Trumpington Street Towson University,York Road 8000 Towson,
Cambridge CB2 1PZ, UK MD 21252, USA
John Clarkson Ann Heylighen
Department of Engineering Department of Architecture
Cambridge Engineering Design Centre Urbanism and Planning
University of Cambridge Katholieke Universiteit Leuven
Trumpington Street Kasteelpark Arenberg 1/2431
Cambridge CB2 1PZ, UK 3001 Leuven, Belgium
Peter Robinson
Computer Laboratory
University of Cambridge
JJ Thomson Avenue, Madingley Road William
Gates Building 15
Cambridge CB3 0FD, UK
ISBN 978-1-4471-2866-3 e-ISBN 978-1-4471-2867-0

DOI 10.1007/978-1-4471-2867-0
Springer London Heidelberg New York Dordrecht
British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library
Library of Congress Control Number: 2012933436
Ó Springer-Verlag London 2012

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material
is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter
developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis
or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive
use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the
provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must
always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright
Clearance Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication, neither
the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may
be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.
Cover Photography: Matt Willox in Design Futures, part of Sheffield Hallam University
Packaging Accessibility Analysis, Dr Alaster Yoxall: Lab4living, Sheffield Hallam Universit
Museum M in Leuven, Belgium designed by Stéphane Beel Architecten: Peter-Willem Vermeersch
Cover design: eStudio Calamar S.L.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Preface
The Cambridge Workshops on Universal Access and Assistive Technology

(CWUAAT) are a series of workshops held at a Cambridge University College
every two years. This volume: “DESIGNING INCLUSIVE SYSTEMS” comes
from the 6th in this series of highly successful events. The workshops are
characterised by a single session running over three days in pleasant surroundings
with delegates from home and abroad staying on site. Feedback suggests that
allowing speakers longer presentation times, carrying discussion on through
sessions into plenaries and shared mealtimes generates a highly cooperative and
creative academic environment that is both enjoyable and informative.
The workshop theme: “Designing inclusion for real-world applications” refers
to the emerging potential and relevance of the latest generations of inclusive design
thinking, tools, techniques and data, to mainstream project applications such as
healthcare and the design of working environments. Inclusive Design Research
involves developing tools and guidance enabling designers to design for the widest
possible population, for a given range of capabilities. In the context of
demographic changes leading to an increasing number of older people, the general
field of inclusive design research strives to relate the capabilities of the population
to the design of artefacts, environments and technology by better characterising the
user and the task demand. Inclusive populations of older people, for example,
contain a greater variation in sensory, cognitive and physical capabilities. These
variations may be co-occurring and rapidly changing leading to a demanding
design environment.
Previous research developments in inclusive design have addressed issues of
matching product and task demand to users’ capabilities in the context of simple
daily living activities or specific products. New research developments are now
extending the scope of the inclusive design approach into real-world applications
by forming interdisciplinary links with systems engineering, industrial product
design, healthcare and medical device design as well as education, policy
development and architecture. This is a necessary stage of research because once
design techniques and materials are fully developed for knowledge transfer, a
v
vi Preface
proving ground is required in real-world application and industry. This proving

ground then tests the impact made by the original research.
As in the previous years, this book contains the best reviewed papers invited for
oral presentation. The papers that have been included were selected by blind peer
review carried out by an international panel of currently active researchers. The
chapters forming the book represent an edited sample of current national and
international research in the fields of inclusive design, universal access, and
assistive and rehabilitative technology.
In the 2012 workshop, as well as the typical Inclusive Design themes of
measuring demand and capability; emergent technologies, and design for inclusion,
there has also been more focus on new themes such as cognitive interaction with
new technologies, architecture, and healthcare. This reflects the newly developing
transdisciplinary perspectives and ongoing research agendas. For example, can
medical and neuroscientific models of thinking impairment be harmonised with
functional descriptions to assist more inclusive design? Is it possible to motivate
older generations to use modern healthcare software by better understanding the
psychology of human motivation? Can we identify and quantify the differences
between designers’ and users’ mental models of a product? In addition,
researchers are increasingly investigating how public policies; both from
governments and international non-governmental organisations, influence inclusive
and accessible design, as well as the usage and adoption of assistive technology by
individuals. Healthcare is a forcing domain: how can we provide architects with
sufficient evidence to enable them to design healthcare buildings that better
anticipate the needs of patients lying in a hospital bed?
For this CWUAAT we have extended the editorial panel to include two
esteemed colleagues, Ann Heylighen from KU Leuven in Belgium, and Jonathan
Lazar from Towson University in the US. This reflects the growing importance of
particular interdisciplinary fields such as inclusive architecture, and public policy
related to inclusive design, to the CWUAAT workshops. It also acknowledges the
substantial international contributions that have been made over the series.
There are five main themes:

I. Designing for the Real-world addresses the application of Inclusive Design
techniques in healthcare, public facilities and services, and hazardous
traditional industries;
II. Measuring Demand and Capabilities looks at ways of measuring
capability-demand relationships for actual tasks, software, devices and
buildings;
III. Designing Cognitive Interaction with Emerging Technologies draws
together a number of threads related to cognition including the alignment
of design and user mental models, motivating older users and unifying
models of cognitive impairment;
IV. Design for Inclusion is a space specifically for design issues in inclusive
design, from sampling through to policy and novel new ways to inform the
designer about inclusive design features;
Preface vii
V. Designing Inclusive Architecture highlights specific cases, such as

inclusive heritage, architecture for dementia and virtual environment tools
for design.
In the tradition of CWUAAT, we have solicited and accepted contributions
over a wide range of topics, both within individual themes and also across the
workshop’s scope. We ultimately hope to generate more interdisciplinary dialogues
based on focused usage cases that can provide the discipline necessary to drive
further novel research, leading to better designs. The aim is to impact industry and
end-users as well governance and public design, thereby effectively reducing
exclusion and difficulty in people’s daily lives and society.
We would like to thank all those authors and contributors who have submitted
to CWUAAT 2012 and to the preparation of this book. Many thanks are also due to
the reviewing members of the Programme Committee who continue to support the
workshop series. Finally, thanks are particularly due to Mari Huhtala and Suzanne
Williams, who both play a key role in bringing the resulting publication to fruition
between final submission and the Workshop itself. As in previous years, we are
grateful to the staff at Fitzwilliam College for their patience and service.
Pat Langdon, John Clarkson, Peter Robinson

Jonathan Lazar and Ann Heylighen
The CWUAAT Editorial Committee

University of Cambridge
March 2012
Contents
List of Contributors …………………………………………………………..xiii
Part I Designing for the Real-world

1. Hospital Reality from a Lying Perspective: Exploring a Sensory
Research Approach
M. Annemans, C. Van Audenhove, H. Vermolen
and A. Heylighen .................................................................................. 3
2. Inclusive Bus Travel - A Psychosocial Approach

F. Nickpour, P.W. Jordan and H. Dong.............................................. 13
3. Safe and Inclusive Design of Equipment Used in the

Minerals Industry
T. Horberry and T. Cooke .................................................................. 23
4. Deploying a Two-player System for Arm Rehabilitation

in Schools
R.J. Holt, A.P.H. Weightman, J.F. Gallagher, N. Preston,
M.C. Levesley, M. Mon-Williams and B. Bhakta ................................ 33
5. Evaluating the Accessibility and Usability of Blogging

Platforms for Blind Users
B. Wentz, M. Cirba, N. Kharal, J. Moran and M. Slate ...................... 43
ix
x Contents
Part II Measuring Demand and Capabilities

6. A Population Perspective on Mobile Phone Related Tasks
M. Bradley, S. Waller, J. Goodman-Deane, I. Hosking, R. Tenneti,
P.M. Langdon and P.J. Clarkson ....................................................... 55
7. How to Use Virtual and Augmented Reality Techniques to

Design Highly Usable Human-machine Interfaces
S. Ceccacci, M. Germani and M. Mengoni ....................................... 65
8. Development and Evaluation of Sonified Weather Maps for

Blind Users
R. Weir, B. Sizemore, H. Henderson, S. Chakraborty
and J. Lazar........................................................................................ 75
9. Achieving Inclusion in Public Spaces: A Shopping Mall Case

Study
Y. Afacan............................................................................................ 85
10. Visibility Prediction Software: Five Factors of Contrast

Perception for People with Vision Impairment in the Real World
H. Dalke, A. Corso, G. Conduit and A. Riaz ...................................... 93
Part III Designing Cognitive Interaction with

Emerging Technologies
11. Intrinsic Motivation and Design of ICT for the Ageing
Population
T.S. Goldhaber, P.M. Langdon and P.J. Clarkson........................... 105
12. A Framework for Studying Cognitive Impairment to

Inform Inclusive Design
E. Jokisuu, P.M. Langdon and P.J. Clarkson................................... 115
13. Interactive Error Correction Using Statistical Language Models

in a Client-server Interface for Editing Mathematical Text
D. Attanayake, G. Hunter, E. Pfluegel and
J. Denholm-Price.............................................................................. 125
14. Understandable by Design: How Can Products be Designed

to Align with User Experience?
A. Mieczakowski, P.M. Langdon, R.H. Bracewell, J.J. Patmore
and P.J. Clarkson............................................................................. 133
Contents xi
Part IV Design for Inclusion

15. Design Advisor: How to Supply Designers with Knowledge
about Inclusion?
E. Zitkus, P.M. Langdon and P.J. Clarkson ..................................... 145
16. From Guinea Pigs to Design Partners: Working with Older

People in ICT Design
R. Edlin-White, S. Cobb, A. Floyde, S. Lewthwaite, J. Wang and
J. Riedel ........................................................................................... 155
17. When Users Cannot be Included in Inclusive Design

R. Herriott ......................................................................................... 165
18. What is Good Design in the Eyes of Older Users?

N. Goddard and C. Nicolle ............................................................... 175
19. Equal Access to Information? Evaluating the Accessibility of

Public Library Web Sites in the State of Maryland
J. Lazar, B. Wentz, C. Akeley, M. Almuhim, S. Barmoy, P. Beavan,
C. Beck, A. Blair, A. Bortz, B. Bradley, M. Carter, D. Crouch,
G. Dehmer, M. Gorman, C. Gregory, E. Lanier, A. McIntee,
R. Nelson Jr., D. Ritgert, R. Rogers Jr., S. Rosenwald, S. Sullivan,
J. Wells, C. Willis, K. Wingo-Jones and T. Yatto ............................. 185
20. Clustering User Data for User Modelling in the GUIDE

Multi-modal Set-top Box
P.M. Langdon and P. Biswas ........................................................... 195
Part V Designing Inclusive Architecture

21. Inclusive Built Heritage as a Matter of Concern: A Field
Experiment
A. Heylighen ..................................................................................... 207
22. Designing a Virtual Environment Framework for Improving

Guidance for the Visually Impaired
S. Kammoun, M.J-M. Macé, B. Oriola and C. Jouffrais ................... 217
23. Spatial Clues for Orientation: Architectural Design Meets

People with Dementia
I. Van Steenwinkel, C. Van Audenhove and A. Heylighen .............. 227
Index of Contributors …………………………………………………..…..237

List of Contributors
Afacan Y., Department of Interior Architecture and Environmental Design,

Faculty of Art, Design and Architecture, Bilkent University, Ankara, Turkey
Akeley C., Department of Computer and Information Sciences, Universal Usability
Laboratory, Towson University, US
Almuhim M., Department of Computer and Information Sciences, Universal
Usability Laboratory, Towson University, US
Annemans M., Department of Architecture, Urbanism and Planning, Katholieke
Universiteit Leuven, Heverlee, Belgium
Attanayake D., Faculty of Computing, Information Systems and Mathematics
(CISM), Kingston University, London, UK
Barmoy S., Department of Computer and Information Sciences, Universal
Beavan P., Department of Computer and Information Sciences, Universal
Beck C., Department of Computer and Information Sciences, Universal Usability
Bhakta B., Academic Department of Rehabilitation Medicine, Faculty of Medicine
and Health, University of Leeds, Leeds, UK
Biswas P., Engineering Design Centre, Department of Engineering, University of
Cambridge, Cambridge, UK
Blair A., Department of Computer and Information Sciences, Universal Usability
Bortz A., Department of Computer and Information Sciences, Universal Usability
Bracewell R.H., Engineering Design Centre, Department of Engineering,
University of Cambridge, Cambridge, UK
xiii
xiv List of Contributors
Bradley B., Department of Computer and Information Sciences, Universal

Bradley M., Engineering Design Centre, Department of Engineering, University of
Carter M., Department of Computer and Information Sciences, Universal
Ceccacci S., Department of Management and Industrial Organisation, Marche
Polytechnic University, Ancona, Italy
Chakraborty S., Department of Computer and Information Sciences, Universal
Cirba M., Division of Business, Management and Technology, Keystone College,
La Plume, PA, US
Clarkson P.J., Engineering Design Centre, Department of Engineering, University
of Cambridge, Cambridge, UK
Cooke T., Minerals Industry Safety and Health Centre, Sustainable Minerals
Institute, The University of Queensland, St Lucia, Australia
Cobb S., Human Factors Research Group, Faculty of Engineering, The University
of Nottingham, Nottingham, UK
Conduit G.J., Department of Condensed Matter Physics, Weizmann Institute of
Science, Rehovot, Israel
Corso A., Design Research Centre, Faculty of Art, Design and Architecture,
Kingston University, UK
Crouch D., Department of Computer and Information Sciences, Universal
Dalke H., Design Research Centre, Faculty of Art, Design and Architecture,
Dehmer G., Department of Computer and Information Sciences, Universal
Denholm-Price J., Faculty of Computing, Information Systems and Mathematics
(CISM), Kingston University, London, UK
Edlin-White R., Human Factors Research Group, Faculty of Engineering, The
University of Nottingham, Nottingham, UK
Elton E., Ergonomics and Safety Research Institute, Loughborough University,
Loughborough, UK
Floyde A., Human Factors Research Group, Faculty of Engineering, The
University of Nottingham, Nottingham, UK
Gallagher J.F., Institute of Engineering Systems and Design, School of
Mechanical Engineering, University of Leeds, Leeds, UK
Germani M., Department of Industrial Engineering and Mathematical Sciences,
Marche Polytechnic University, Ancona, Italy
Goddard N., Ergonomics and Safety Research Institute, Loughborough
University, Loughborough, UK
Goldhaber T.S., Engineering Design Centre, Department of Engineering,
Cambridge University, Cambridge, UK
Goodman-Deane J., Engineering Design Centre, Department of Engineering,
Cambridge University, Cambridge, UK
List of Contributors xv
Gorman C., Department of Computer and Information Sciences, Universal

Gregory C., Department of Computer and Information Sciences, Universal
Henderson H., Department of Computer and Information Sciences, Universal
Herriott R., Aarhus School of Architecture, Aarhus, Denmark
Heylighen A., Department of Architecture, Urbanism and Planning, Katholieke
Universiteit Leuven, Leuven, Belgium
Holt R.J., Institute of Engineering Systems and Design, School of Mechanical
Engineering, University of Leeds, Leeds, UK
Horberry T., Minerals Industry Safety and Health Centre, Sustainable Minerals
Institute, The University of Queensland, St Lucia, Australia
Hosking I., Engineering Design Centre, Department of Engineering, University of
Hunter G., Faculty of Computing, Information Systems and Mathematics (CISM),
Kingston University, London, UK
Jokisuu E., Engineering Design Centre, Department of Engineering, University of
Jouffrais C., IRIT-CNRS, University of Toulouse, Toulouse, France
Kammoun S., IRIT-CNRS, University of Toulouse, Toulouse, France
Kharal N., Division of Business, Management and Technology, Keystone College,
La Plume, PA, US
Langdon P.M., Engineering Design Centre, Department of Engineering,
Lanier E., Department of Computer and Information Sciences, Universal Usability
Lazar J., Department of Computer and Information Sciences, Universal Usability
Levesley M.C., Institute of Engineering Systems and Design, School of
Mechanical Engineering, University of Leeds, Leeds, UK
Lewthwaite S., Department of American and Canadian Studies, University of
Nottingham, Nottingham, UK
Macé M.J-M., IRIT-CNRS, University of Toulouse, Toulouse, France
Mengoni M., Design Tools and Methods Group, Marche Polytechnic University,
Ancona, Italy
Mieczakowski A., Engineering Design Centre, Department of Engineering,
McIntee A., Department of Computer and Information Sciences, Universal
Mon-Williams M., Institute of Psychological Sciences, University of Leeds,
Leeds, UK
Moran J., Division of Business, Management and Technology, Keystone College,
La Plume, PA, US
Nelson Jr.D., Department of Computer and Information Sciences, Universal
xvi List of Contributors
Nickpour F., Inclusive Design Research Group, School of Engineering and

Design, Brunel University, West London, UK
Nicolle C., Ergonomics and Safety Research Institute, Loughborough University,
Loughborough, UK
Oriola B., IRIT-CNRS, University of Toulouse, Toulouse, France
Patmore J.J., Member of Pembroke College, University of Cambridge,
Cambridge, UK
Pfluegel E., Faculty of Computing, Information Systems and Mathematics (CISM),
Kingston University, London, UK
Preston N., Academic Department of Rehabilitation Medicine, University of
Leeds, Leeds, UK
Riaz A., Design Research Centre, Faculty of Art, Design and Architecture,
Riedel J., University of Nottingham Business School, University of Nottingham,
Nottingham, UK
Ritgert D., Department of Computer and Information Sciences, Universal
Rogers Jr.R., Department of Computer and Information Sciences, Universal
Rosenwald S., Department of Computer and Information Sciences, Universal
Sizemore B., Department of Computer and Information Sciences, Universal
Slate M., Division of Business, Management and Technology, Keystone College,
La Plume, PA, US
Sullivan S., Department of Computer and Information Sciences, Universal
Tenneti R., School of Primary, Aboriginal and Rural Health Care, Faculty of
Medicine, Dentistry and Health Sciences, The University of Western Australia,
Crawley, Western Australia
Van Audenhove C., Faculty of Medicine, Katholieke Universiteit Leuven,
Leuven, Belgium
Van Steenwinkel I., Department of Architecture, Urbanism and Planning,
Katholieke Universiteit Leuven, Leuven, Belgium
Vermolen H., Osararchitects nv, Antwerpen, Belgium
Waller S.D., Engineering Design Centre, Department of Engineering, Cambridge
University, Cambridge, UK
Wang J., University of Nottingham Business School, University of Nottingham,
Nottingham, UK
Weightman A.P.H., School of Mechanical Engineering, University of Leeds,
Leeds, UK
Weir R., Department of Computer and Information Sciences, Universal Usability
Wells J., Department of Computer and Information Sciences, Universal Usability
Wentz B., Department of Computer Science and Information Technology,
Frostburg State University, Frostburg, MD, US
List of Contributors xvii
Willis C., Department of Computer and Information Sciences, Universal Usability

Wingo-Jones K., Department of Computer and Information Sciences, Universal
Yatto T., Department of Computer and Information Sciences, Universal Usability
Zitkus E., Engineering Design Centre, Department of Engineering, Cambridge
University, Cambridge, UK
Part I
Designing for the Real-world

Chapter 1
Hospital Reality from a Lying Perspective:

Exploring a Sensory Research Approach
M. Annemans, C. Van Audenhove, H. Vermolen and
A. Heylighen
1.1 Real Buildings, Real Experiences

Despite many efforts by healthcare providers, for most people a hospital stay is
rarely a pleasant experience. The hospital building as such is part of this
perception. Moreover, the specific situation of a hospital stay is largely determined
by the material reality of the organisation. Studies on hospital environments tend to
single out one particular aspect, e.g. the view through the window, or presence of
green (Ulrich 1984a, 1984b) and try to prove its clinical outcome. Yet they fail to
translate their results to the design of real-life settings (Rubin et al., 1998, Cbz
2008). Moreover, the influence of patients’ peculiar perspective, i.e. lying in a
hospital bed, on the way they experience the reality of the hospital is largely under
researched.
The overall objective of our research is therefore to investigate what spatial
aspects influence patients’ well-being in a hospital setting through an improved
understanding of people’s spatial experience from a lying perspective. By
developing a better understanding of the relationship between the patient, the
objects that feature in his or her hospital life, especially the bed and the building,
we hope to enable architects to design buildings that add to the healing character of
the hospital environment. Ultimately, we aim to provide architects with sufficient
evidence to design healthcare buildings that can better anticipate the needs of
patients and other users.
Since our research aims to gain insight in how patients experience a hospital
from a lying perspective, we need a manner to make their spatial experience more
explicit. Therefore, we explored different ways for patients to document their
spatial experience. In this paper, we report on a pilot study which explores how
three patients with a very different profile each deal with this task in their own
way. The empirical material collected is not only very rich in itself, but also
inspires the participating patients to talk about those aspects of the building that
P. Langdon et al. (eds.), Designing Inclusive Systems, 3

DOI: 10.1007/978-1-4471-2867-0_1, © Springer-Verlag London 2012
4 Annemans et al.
affect them most. Certain themes frequently return in the conversations, yet the
goal at this stage in our research is not so much to obtain valid information, but
rather to explore the possibilities of using participant collected material to facilitate
the interviews.
1.2 Sensory Reality

Research about healing or wholesome environments in health care settings is
mostly conducted in the field of evidence based design (EBD). This notion found
its origin in the analogy with other evidence-based approaches to research and
practice. When it comes to buildings for health care especially, evidence based
medicine was a source of inspiration. In EBD studies evidence for the healing
outcome of building aspects is being collected (Ulrich et al., 2004). Mostly
individual aspects are investigated and clinical output is highly valued (Rubin et
al., 1998). However, almost all reports that collect and evaluate the results of the
conducted studies have to conclude that the evidence is not ready to be transposed
to the real world. To start with, the settings in which the trials are carried out single
out individual aspects and, as such, are not representative of real-world hospital
settings. Secondly the studies do not provide insight in how the different aspects
relate to each other (Rubin et al., 1998; Van den Berg, 2005; Cbz, 2008).
However, a wholesome environment involves more than the measurable
aspects that withstand the evaluation of EBD studies. After all, architecture is
experienced through the senses (Rasmussen, 1964; Pallasmaa, 2005), so how a
place feels, smells, sounds, and looks defines our impression of it. In this multi-
sensory experience of the environment, time, movement and activities play a major
role. The senses are indeed a key factor in the experience of everyday reality
(Ingold, 2000). If we are to develop an articulate understanding of patients’ actual
experience from a lying perspective, as is the aim of our research, we thus need to
collect firsthand information about what they feel, sense, and think during their
hospital stay, not only while in their room, but also while travelling through the
hospital building.
1.3 Insight/Inside Spatial Experience

In studying patients’ spatial experience from a lying perspective, we explore and
develop a research methodology that respects the interrelatedness of how they
experience the reality of things in the hospital, addressing different aspects of the
built environment as well as their complex interaction. To obtain a broad spectrum
of empirical material, participants should be challenged to interrogate all their
senses. The method of collecting material should also appeal to a very diverse
group of patients within the hospital, making them feel at ease with their
participation. Only when these criteria are met, will the material be rich enough to
gain a nuanced understanding of the complex reality of everyday hospital
experience.
Hospital Reality from a Lying Perspective 5
In the first months of our research we conducted interviews with various actors
in the field (medical staff, patients, technical directors of hospitals, hospital
architects) to obtain a profound understanding of the meaning of the bed in the
hospital (Annemans et al., 2011). Each interviewee shed light on the topic from
his/her specific perspective. This resulted in a good understanding of the
complexity of hospital design. Although our research focuses on the experience of
(lying) patients, this background makes us aware of the importance of reflecting on
and taking into account certain undeniable technical and organisational realities.
During these interviews it became clear that people find it hard to talk about their
spatial experience, especially those unfamiliar with the design and construction
process. This lack of conversation skills about the research topic makes it hard to
obtain the rich empirical material we are looking for. Part of the difficulty can be
explained by a lack of vocabulary when talking about space (Franck and von
Sommaruga Howard, 2009). Additionally, due to the complexity of experience, it
is not easy to ensure that it is explored throughout the entire range of its various
articulations (Throop, 2003). These obstacles force us to look for a suitable
research approach that can give us access to patients’ personal ways of
experiencing a hospital environment.
For this reason we decided to explore alternative interviewing techniques.
Aware of the difficulty for patients of expressing their spatial experience and given
the explorative phase of the research, we opted to ask the participating patients
only one, very broad question: “Can you document the hospital experience from a
lying perspective?” After a short introduction patients are then invited to document
their spatial experience in any way that pleases them. They are provided with pens,
pencils, note block, drawing paper and a camera (with the possibility of recording
sound and movies), or they can use their own camera. Afterwards the material
produced during this process is used to facilitate a semi-structured interview with
the participating patient about the spatial qualities of the building.
As already pointed out, we experience the built environment through all our
senses. For this reason we looked into the use of sensory methods. Visual
techniques like photo-elicitation where people are interviewed using photographs
are not new (Collier, 1967). Over the years the technique has gained popularity in
various fields such as visual ethnography, visual anthropology, visual sociology,
and visual cultural studies (Pink, 2007). A sensory method does not guarantee
access to sensory realms; for example, the visual is not necessarily best accessed
by a visual method (Mason and Davies, 2009). Nevertheless, introducing
photography can be seen as a first contribution to a more ‘sensually complete’
methodology (Warren, 2002). In order to capture the real-life experience instead of
a fictive reality constructed by the researcher, however, we shifted our focus
beyond photo-elicitation to photo-production (Radley, 2010). In line with the work
of other researchers (Herssens and Heylighen, 2009; Radley, 2010), our interest
lies not so much in the meaning of pictures, we want to gain a good understanding
not only of what has been made visible, but also why, and how.
Indeed, how images are recorded may also tell us something about the hospital
experience of the participating patients. As taking pictures in hospitals is not
always appropriate or allowed (Radley and Taylor, 2003), we did not provide the
patients with a camera only, and limit their way of expression to photographs, but
6 Annemans et al.
also gave them the opportunity to take notes or make a drawing. This approach
yields richer empirical material, which initiates different topics in the interviews.
In the elicitation process, there is no reason why the drawings and plans should be
treated any differently to the photographs (Harper, 2002).
1.4 “Talking” Patients, “Speaking” Hospitals

1.4.1 Who is Talking
To explore the methodology’s possibilities and limitations we try it out with three
very different, pragmatically chosen patients. Because of the explorative stage of
the research the validity and generalisability of the obtained results is subordinate
to the test of the methodology and the participants’ engagement. The three of them
are in the hospital for different reasons while documenting their stay. Although all
are provided with the same equipment, they choose very diverse ways to document
their experience.
There is Mrs A, who stays a few days in the hospital for rather serious surgery.
Through photographs she elaborately documents her experiences in two rooms, a
double room before the operation, a single room afterwards. Despite her
commitment to the task, she admits she does not feel comfortable or able to take
the camera into the hallway or to treatment.
Mrs B has a long history of hospital visits, and participates in this study when
going for a check-up in a hospital where she has been a regular patient. Since she is
not lying herself while travelling through the hospital this time, she interprets the
question as illustrating what she remembers from previous visits and uses it to
signal points of improvement in the treatment of patients. Depending on what she
feels is appropriate, she switches between taking pictures and drawing sketches of
what she observes. Even a brief story emerges.
Mr C is on dialysis, making him visit the same part of the hospital several times
a week. Taking pictures he considers redundant, as the researcher can see
everything herself when conducting the interview. However, he does have some
ideas about how the department should be rearranged, improving most of the
disadvantages he experiences today. He sketches a plan of the current situation to
explain to the researcher how it is and re-sketches it to illustrate how he would
alter it.
1.4.2 How Patients “Talk”

How patients choose to document their experience tells us a lot about how they
experience certain situations. The pencils and papers, for instance, were initially
introduced as a pragmatic solution to the problem that taking pictures in hospitals
is not always appropriate or allowed. However, they turn out to provide insights
that would have stayed undiscovered had only a camera been used.
As one patient points out, sketching an awkward situation afterwards is much less
confronting for the people involved than taking a picture as it unfolds. Another
patient does not see the point of taking pictures of a building that the researcher can
observe herself, however he does have some suggestions on how a future hospital
setting should be designed. Whereas, for him, a photograph just shows an existing
situation, drawing provides the opportunity to manipulate reality.
Since we use the recorded images as a base for interviews, it does not come as a
surprise that also this kind of material serves as an interesting source of information
about how the hospital building is perceived, or which aspects are appreciated or
considered problematic. Often just starting the conversation is enough to divert to
aspects not, or only slightly, related to the collected material. As if enabling
participants to start the conversation from their viewpoint is enough to open the door
to unexpected insights.
1.4.3 What Hospitals “Say”

Through the material collected by the participants, the hospitals tell us both about
their organisation and about the building and its interior. The collected images and
accompanying narratives give us a good overview of what kind of material can be
expected. Obviously there are the images that literally represent visual elements in
the environment, such as the ceiling. However mostly, here too, the explanation why
the image is recorded reveals supplementary useful insights. In addition some images
are not taken for what they show but have an iconographic meaning, telling us about
other sensory experiences than the visual. Finally, the produced images can be a
representation of what could be, giving the participating patients a voice to express
how they would change the spatial situation they are in. Sometimes the same picture
can be placed in different categories because of the multi-layered explanation that
comes with it.
Pictures of the windows or the ceiling are fairly straightforward in what they
indicate. Fresh air and a view of the outside world are generally considered important
aspects of a healing environment (Rubin et al., 1998; Devlin and Arneill, 2003;
Dijkstra et al., 2006; Cbz, 2008). It is thus no surprise that one who wants to discuss
these topics pictures windows. When lying in a hospital bed your perspective
changes. Lying or sitting in bed has a great impact on what, or whether, you see
through the window.
Figure 1.1. Difference between what you see through the window when lying down versus
sitting in bed
8 Annemans et al.
The iconographic meaning can only be discovered when both the image and the
narrative are combined. Indeed, the same object can be pictured for a completely
different reason. One of the pictures made by Mrs A shows a window as well.
However, this window does not represent fresh air or daylight. The window in the
picture is located in an inner wall around the atrium accommodating the cafeteria.
Through the window the sound of chattering people and laughing children enters
the room. Although Mrs A describes this as a pleasant sound when in a different
setting, she finds it quite disturbing while being in the hospital and trying to
recover from surgery.
Fairly banal objects can be icons of less tangible aspects that have a strong
impact on the experience of the hospital. Mrs A photographs the little table and the
chairs in her room while having visitors. Because of the table’s presence, the
people in the chairs do not seem to be visiting in a hospital so much as just being at
home. As she puts it:
‘I think it has a more homelike atmosphere when people can be sitting on a chair, at
a table, where you can put something on, than when you would just have a row of
chairs, then it would feel like they were watching me.’
(Mrs A, interview transcript)
Figure 1.2. Iconographic images: left: Window photographed to illustrate the sound that
enters through it; right: The presence of a table changes the perception of visitors on the
chairs.
Mrs B made a set of two photos and a drawing, picturing the transportation of
a(n unknown) patient in bed. In this trilogy she combines a reflection on her own
experiences while transported through the hospital in a bed and the observations
from the bed.
Figure 1.3. Nurse connecting a patient’s bed to a wagon, a view of the ceiling while
travelling through the hospital, setup of how the bed is transported through the hospital
The first picture shows the hallway in the basement of the hospital. Mrs B
mainly wants to point out how she felt when she was transported through it during
an earlier stay. It is a very functional hallway, used as a storage space for carts,
bicycles, obsolete beds and so on. It seems as if no one ever thought about the fact
that patients who have to go to the nuclear department are transported through it.
As Mrs B explains, “the lighting is not pleasant, and all the stuff that is stored
there makes you feel uncomfortable”. The nurse in the picture is attaching a bed to
a cart, used to pull the beds when large distances have to be covered. Mrs B
reflects on how the patient in the bed must feel, being handled like that. Since she
did not feel comfortable taking a picture while the patient could see her, she drew
the setup, explaining how the bed is attached to the cart, how the patient does not
have any contact with the nurse involved, and what s/he must perceive and feel
while being towed like that. To illustrate that, Mrs B also took a picture of the
ceiling a little further down the hall and concluded that seeing all those pipes, and
the rags in between, is not how a disconcerted patient is comforted. The dust
between the technical equipment on the ceiling makes her wonder about hospital
hygiene. Although these pictures might seem a little banal at first, how they come
together and are used as a backbone for the narrative provides new opportunities to
talk about spatial experience. We do not just learn about Mrs B’s experience while
visiting the hospital this time, but are also provided with a reflection on previous
visits and when and why you feel most vulnerable as a patient, which she uses to
explain her reluctance to take a picture. Even movement and time are touched upon
during the conversation. The length of the travel through the hall, how the patient
must undergo the transportation not knowing where he/she is going, it all adds to
10 Annemans et al.
the experience of the building without being necessarily related to visual

perception.
In response to the simple question we asked the patients to also express their
view on how the building could be improved. For Mr C drawing a plan of the ward
gives him the opportunity to point out the shortcomings of the current setup, while
by sketching how he would organise it, he brings up spatial aspects that he thinks
are important. More acoustic barriers, more visual transparency, and a reduction of
the passage to create a quiet atmosphere are only some of the aspects mentioned.
Figure 1.4. Left: existing situation, right: how Mr C would organise the ward
Suggesting alterations to the hospital layout is not restricted to drawing plans.

Mrs A documents how she improved her privacy by opening a second door to
block the view from the hallway. When both her door and the one at the other side
of the hall were fully opened, she could see right into the other room. Since she did
not want to close her door completely, preferring some connection with the rest of
the ward, she opened the door of the bathroom to block the view. The conversation
about these pictures first focuses on the privacy but soon broadens to interactions
with other patients and staff, feelings of safety and even how bed transport could
be improved so there would be fewer obstacles on the way.
Figure 1.5. Left: open doors, giving a view into the room at the other side of the hall, right:
opening the bathroom door can create a second barrier without closing the door of the room
1.5 Conclusions and Future Work

Developing a better understanding of how patients experience the complex reality
of a hospital building from a lying perspective, asks for a sensitive methodological
approach that addresses all senses and facilitates reflection on experience.
Although we do not wish to push forward a single method, the technique explored
so far seems to fulfil the requirements. By asking patients to document their spatial
experience and providing them with multiple media to do so, we gain insights from
the inside out, not only from the material they collect, but also from why and how
they collect it. Using the illustrations generated by the participants themselves
clearly adds an extra dimension to the interviews. Both the interviews and the
findings are deepened. Even though the material collected is mostly visual, it can
also illustrate auditory or haptic qualities, like the window through which a lot of
noise enters. The images, complemented by the accompanying narratives, serve as
a source of information about all senses, movement and time, illustrating the
intangibility of sensory experience. Overall, this alternative way of questioning
people definitely provides additional value given the aim of our research.
Given the promising results of this exploratory phase, the approach will be
further elaborated in future phases into a more generally applicable methodology
that can, at least partially, bypass the difficulty of articulating spatial experience.
The motivation of the participating patients is of crucial importance to the success
of the approach. As such, finding patients willing to participate is a key concern.
Although part of the strength of the approach lies in its flexibility, a balance should
be found between freedom and guidance. In future research, when the key concern
is the outcome of the interviews and not the methodology followed, the selection of
the settings and participants will need to be more representative to obtain more
valid and transferable results.
When thinking about the translation of the research results to inform architects’
design process, we might even consider using the empirical material in its original
form. This would imply that real data, directly from the patients, are used by those
designing environments for them. Especially in health care buildings, the highly
complex reality forms an important step for designers to acknowledge the needs of
the primary users. Since the evidence provided by EBD studies conducted so far is
not ready to be converted to real-life applications yet, it is definitely worth
investigating whether different empirical material can be used to inform the design
process.
1.6 Acknowledgements
Margo Annemans’ research is funded by a PhD grant from the Baekeland program
from the Institute for the Promotion of Innovation through Science and Technology
in Flanders (IWT-Vlaanderen). The Baekeland program gives researchers the
opportunity to complete a PhD in close collaboration with the industry, in this case
Osar Architects nv provided this opportunity. Ann Heylighen received support
form the European Research Council under the European Community’s Seventh
12 Annemans et al.
Framework Programme (FP7/2007-2013)/ERC grant agreement No. 201673. The

authors thank the participating patients for sharing their time and insights.
1.7 References
Annemans M, Van Audenhove C, Vermolen H, Heylighen A (2011) Lying architecture:
Experiencing space from a hospital bed. In: Proceedings of the 1st International
Conference Exploring the Multi-dimensions of Well-being (Well-being 2011),
Birmingham, UK
Cbz (2008) Kwaliteit van de fysieke zorgomgeving, stand van zaken onderzoek
omgevingsvariabelen en de effecten op de (zieke) mens, College bouw zorginstellingen,
Utrecht, The Netherlands
Collier J (1967) Visual anthropology: Photography as a research method, Holt Rinehart and
Winston, NY, US
Devlin AS, Arneill AB (2003) Health care environments and patient outcomes: A review of
the literature. Environment and Behavior, 35(5): 665-694
Dijkstra K, Pieterse M, Pruyn A (2006) Physical environmental stimuli that turn healthcare
facilities into healing environments through psychologically mediated effects: Systematic
review. Journal of Advanced Nursing, 56(2): 166-181
Franck K, von Sommaruga Howard T (2010) Design through dialogue: A guide for clients
and architects, Wiley, Chichester, UK
Harper D (2002) Talking about pictures: A case for photo elicitation. Visual Studies, 17(1):
13-26
Herssens J, Heylighen A (2009) A lens into the haptic world. In: Proceedings of the
International Conference on Inclusive Design and Communications (INCLUDE 2009),
London, UK
Ingold T (2000) The perception of the environment: essays on livelihood, dwelling and skill,
Routledge, London, UK
Mason J, Davies K (2009) Coming to our senses? A critical approach to sensory
methodology. Qualitative Research, 9(5): 587-603
Pallasmaa J (2005) The eyes of the skin: Architecture and the senses, John Wiley & Sons,
NY, US
Pink S (2007) Doing visual ethnography: Images, media, and representation in research, 2nd
edn. Sage Publications, London, UK
Radley A (2010) What people do with pictures. Visual Studies, 25(3): 268-279
Radley A, Taylor D (2003) Images of recovery: A photo-elicitation study on the hospital
ward. Qualitative Health Research, 13(1): 77-99
Rasmussen SE (1964) Experiencing architecture, MIT Press, Cambridge, MA, US
Rubin HR, Owens AJ, Golden G (1998) An investigation to determine whether the built
environment affects patients’ medical outcomes, Center for Health Design, Martinez, CA,
US
Throop CJ (2003) Articulating experience. Anthropological Theory, 3(2): 219-241
Ulrich RS, Zimring C, Joseph A, Quan X, Choudhary R (2004) The role of the physical
environment in the hospital of the 21st century: A once-in-a-lifetime opportunity, Centre
for Health Design, Concord, CA, US
Van den Berg AE (2005) Health impacts of healing environments: A review of evidence
for benefits of nature, daylight, fresh air, and quiet in healthcare settings. University
Hospital Groningen, Groningen, The Netherlands
Warren S (2002) ‘Show me how it feels to work here’: Using photography to research
organisational aesthetics. Ephemera Critical Dialogues on Organisations, 2(3): 224-245
Chapter 2
Inclusive Bus Travel - A Psychosocial

Approach
F. Nickpour, P.W. Jordan and H. Dong
2.1 Introduction
2.1.1 Public Transport - The New Climate
Public transport is facing major challenges in the current economic and social
climate; a considerable rise in demand for public transport and an ageing
population that is mainly dependant on public transport and is increasingly in need
of specialised and door-to-door services. The above challenges double when one
considers the raised public awareness and the pressure from user organisations to
improve the equality and quality of public transport for all.
Public transport providers need to respond to increasing demand for service
provision, both in terms of volume and diversity of service users. Transport for
London (TfL), a major public transport provider in UK, is currently facing over-
subscribed door-to-door services and an increasing demand for accessible and
usable public transport by conventionally marginalised groups such as older people
and people with disabilities. Issues of accessibility, reliability and quality of
service are key indicators that are sometimes in conflict and need to be revisited.
There is a need to keep the quality of service consistent and at the same time
redefine and prioritise the areas of focus and improvement.
2.1.2 Public Bus Services

Buses will continue to be - probably for many years - the main and only form of
public transport that can be accessible to almost all (London TravelWatch, 2010).
There is also evidence that bus services are often more frequently used by
disadvantaged or vulnerable sections of society, therefore poor performance is
more likely to impact on these groups (London TravelWatch, 2009). Thus, the bus

14 Nickpour et al.
service proves to be the single most powerful transport tool in terms of inclusivity
and equality potential and provision in a mega-city like London.
There have been great improvements in terms of making buses fully accessible.
In London, all buses are now low-floor vehicles and have a space for one
wheelchair (Transport for London, 2011). However, an ‘accessible bus’ does not
necessarily guarantee an ‘accessible bus service’. An accessible bus service
requires not only an accessible bus and an accessible bus stop but also an empathic
well-trained driver and a user-friendly environment. As well as improving
inclusivity, making local bus services more accessible brings wider benefits
including facilitating social inclusion in the local community, making bus travel
easier and more pleasurable for every member of the local community and
reducing the need for dedicated services (e.g. Dial a Ride) which are not cost-
effective.
2.1.3 The Project

Commissioned by Transport for London and one local London borough, a research
project was conducted in order to address issues associated with bus travel in
London. The aim of the project was to produce recommendations for improving
the accessibility of bus travel through investigating barriers to a diverse range of
people using (or not using) public buses and what makes a journey either pleasant
or unpleasant. A variety of approaches and techniques were used in order to
understand the barriers to accessibility and inclusivity and how these could be
overcome. The research project aimed to assess and improve the accessibility of
public buses through a holistic and comprehensive service-oriented approach,
focusing on an accessible bus service as a whole rather than focusing on segments
of the whole service such as bus or bus-stop.
2.1.3.1 Bus Service - Key Stakeholders

Broadly, with respect to bus services, three major stakeholders were defined:
• Service user - mobility challenged people
• Service provider - bus drivers
• Service operator - bus companies
Addressing accessibility and inclusivity issues, the project focused on mobility
challenged people as the critical bus service users. For the purpose of this project, a
mobility-challenged person was defined:
‘A mobility challenged person is someone whose mobility has been challenged due
to age, physical or mental impairment, or an external physical condition; each of
the above could have substantial and long-term adverse effect on the person’s
ability to use public transport.’
(Nickpour and Jordan, 2011)
This definition includes, but is not limited to, wheelchair users and those with
other impairments that affect mobility. Other major groups with other mobility
Inclusive Bus Travel - A Psychosocial Approach 15
restrictions that may make it more difficult to use public transport are: older
people, blind or visually-impaired people, deaf people or people with hearing
difficulties, those with learning difficulties or social phobias, and guardians with
buggies.
2.1.3.2 Bus Service - Stakeholder Issues

Key issues concerning each stakeholder included:
• For bus passengers: Positive experience from start to finish - every stage of
the journey should be efficient, enjoyable and smooth, and the user should
be and feel safe at all times.
• For bus drivers: Pleasant working environment - drivers should be treated
politely and respectfully by all passengers. They should be equipped with
the skills needed to carry out all aspects of their duties competently and
receive the full support of both bus users and their employers in doing so.
• For bus operators: Profitable business - operators should be encouraged
and enabled to fulfil the service requirements against suitable performance
targets in a manner which is commercially viable.
2.2 Methodology and Methods

2.2.1 Methodology
The research project followed a combined primary and secondary research
methodology, with a heavy focus on primary research conducted through a diverse
range of field research methods. A major focus for the project was consultation
with people who had a wide range of mobility challenges. Many other stakeholders
were also included in the consultation process. This included bus drivers and
representatives from bus operating companies, TfL, police and advocacy groups
representing mobility-challenged people.
In addition to this consultation process, members of the project team gained
first-hand experience of some of the issues faced by mobility-challenged people by
taking bus trips while using wheelchairs. Information was also collected through
observing mobility-challenged people travelling on buses and asking mobility-
challenged residents of London Borough of Hillingdon - where the study was
conducted - to take bus journeys and report their experiences.
2.2.2 Methods
A wide range of methods were used in order to collect first-hand information
regarding the existing barriers and issues regarding accessibility and inclusiveness
of bus services. All primary research was undertaken in the local London borough.
In some cases, similar services were observed in other London boroughs as well.
Due to space limitation, specific details in terms of participants’ process of
16 Nickpour et al.
selection, age, demographics, position, etc. are not included here. More detail on
the above is provided in a technical report (Nickpour and Jordan, 2011).
2.2.2.1 Focus Groups

Three focus group sessions with different focuses were run in order to provide a
holistic understanding of the existing issues. Each session focused on one
stakeholder group. Firstly, a focus group session was held with nine representatives
of service providers and a cross-section of other stakeholders aiming to look at
organisational and big-picture issues. The participants included representatives
from TfL, the local Council, bus companies, Dial a Ride, Age UK, Metropolitan
Police, Hillingdon Community Transport and Access and Mobility Forum. Then,
one session was held with a diverse group of service users with a focus on
mobility-challenged passengers. This included nine participants; one blind person,
one person with learning difficulties, one wheelchair user and six older people.
Finally, a session was held with service non-users including seven mobility-
challenged members of the public who did not use currently public buses for a
variety of reasons. These included previous negative experience with using public
buses and lack of trust and confidence in the service.
2.2.2.2 Access Audits

Two sets of access audits were planned and carried out. The emphasis was on both
immersion (Moore and Conn, 1985) and direct observation (Dray, 1997). The first
series of audits included eight local bus journeys and were carried out by the
project research team, role-playing by using a wheelchair, aiming to look at
specific mobility issues. Each observation session was attended by two members of
the research team. The second series of access audits were carried out by a diverse
group including five local participants with mobility impairments. Participants
included one male older person aged 72, and two wheelchair users, one with an
electric wheelchair and one with a normal wheelchair. Also, one person with
learning difficulty aged 21 and one blind person aged 42 carried out the access
audits. All audit sessions were documented through various applicable audio,
visual and textual formats.
2.2.2.3 Interviews and Meetings

A number of meetings and interviews were held with individuals from various
organisations and groups in order to look into a number of issues in more detail.
Altogether, five interview sessions were held; these included interviews with three
bus drivers, meetings with Hillingdon Community Transport general manager, the
accessibility officer of Hillingdon Council, two officers from the Disablement
Association of Hillingdon and six members of the local Youth Council.
2.2.2.4 Observations
Two major observation sessions were held. One session focused on special services
aimed at mobility-challenged passengers; the project team spent a day working
with the Dial-a-Ride service that provided door to door transport for mobility-
challenged people. Another observation session took place at Bus Mentoring Day -
a training day aimed at helping those who assist mobility challenged people with
their travels.
2.2.2.5 Literature Review

The literature review drew on a number of sources, reports and documents
including reports by the Disabled Persons Transport Advisory Committee
(DPTAC), Direct Gov, The Department of Transport and London TravelWatch.
The main source for the literature review was the new report by the Greater
London Authority (GLA), titled “Accessibility of Transport” (GLA, 2010) which
looked at the accessibility of all public transport within the capital including buses.
The report drew on inputs from a wide variety of advocacy groups representing
mobility challenged people as well as on a wide array of statistics quantifying
accessibility of buses and other modes of transport.
2.3 Findings
Based on the access audits conducted, the journey was broken down into the stages
shown in Figure 2.1.
Figure 2.1. Key stages of a bus journey
The findings are presented under three key categories; physical, psychosocial
and operational issues. Due to the length available for this paper, only a summary
of findings is presented here. Detailed breakdown and analysis of findings can be
found in the ‘Inclusive Bus Travel in Hillingdon: Assessing Accessibility’ report
(Nickpour and Jordan, 2011).
2.3.1 Physical Issues
From a physical accessibility point of view, users tended to find the most
problematic part of the journey was getting from home to the bus stop and getting
from the bus to their final destination. Examples of problems here included: narrow
pavements, loose paving stones, steep roads and difficult crossings. There were
also accessibility difficulties at some bus stops - for example, the positioning of
litter bins and other street furniture sometimes made deploying and using the ramp
somewhat inconvenient.
18 Nickpour et al.
However, despite such difficulties, it was possible for mobility challenged

people to board the bus at all of the stops examined in the audit. Improvements in
the design of buses meant that, in general, once the user had reached the stop, the
bus could be accessed OK and the on-board part of the journey completed.
2.3.2 Psychosocial Issues
Various observational and immersive methods used also uncovered a number of

other difficulties - mostly psychological and social - that users faced. These
included:
2.3.2.1 Uncertainties
There were many aspects to this including uncertainties as to whether users would
be able to get on and off the bus OK, whether they would have a long wait at the
stop and whether their interactions with others would be positive.
2.3.2.2 Overcrowding
The start and end of the school day are times when the bus gets particularly
crowded. This can sometimes mean that the bus is too crowded to let a wheelchair
on. Even if it is possible to board, overcrowding can make it difficult for
wheelchair users to get to the wheelchair bay and to move their chair into the
proper position within it. Overcrowding is becoming an increasingly problematic
issue as more and more people are using buses. This is due in part to the difficult
economic conditions that we have had recently (bus travel tends to increase in
times of financial hardship) and in part to the issuing of free bus passes to
schoolchildren and older people.
2.3.2.3 Negative Experiences with Drivers

Many users had also mentioned that they had had problems with the drivers. This
could be because of inconsiderate driving - for example pulling away too quickly -
or because they were perceived as having an unfriendly or surly attitude towards
the user. Indeed, during the access audits there were a number of incidents of
drivers not stopping at bus stops when they saw a wheelchair user waiting to get
on. Bus drivers mentioned that there were often problems with ramps failing to
deploy and cited this as a reason why they could not always pick up wheelchair
users.
2.3.2.4 Negative Behaviour of Other Passengers

A number of participants reported being annoyed or intimidated by the behaviour
of other passengers. In particular they mentioned teenagers who they said could be
very loud and often used foul language. A number of participants also mentioned
that they also found it annoying when people had loud conversations on mobile
phones or played music so loudly that it could be heard through their headphones.
The behaviour of other passengers when getting on and off the bus was also a
source of annoyance and intimidation. In particular they mentioned pushing and
shoving and people not waiting their turn in the queue. Other users had reported
that they are wary of using buses in the evening or night because of the risk of
encountering drunk or threatening people.
2.3.2.5 Off-putting Stories

In some cases, participants were put off using the bus because of stories they heard
about other people having bad experiences, in particular stories of violent or
frightening incidents. These stories may have been told to them by friends or they
may have read or heard about them in the media.
2.3.3 Operational Issues

An issue that may be a contributory factor is the key performance indicators (KPIs)
used to measure the performance of the bus operators. Currently, emphasis is
mostly on reliability - that has to do with timeliness of the bus service. There are
no measures in place to monitor either the number of mobility challenged people
using buses or the quality of their experience as one performance indicator.
It was observed that it can take some time for a mobility challenged person,
such as a wheelchair user, to board the bus. This may lead to the bus running
behind schedule with the consequence that it affects reliability. As reliability is the
basis on which the bus companies are judged and the pressure is for them to run on
time, drivers sometimes feel unenthusiastic about picking up mobility challenged
passengers and hence may have a hostile attitude towards mobility challenged
them or may try to avoid picking them up altogether.
2.4 Discussion
2.4.1 Physical Versus Psychosocial Issues
Overall the research suggested that good progress had been made in terms of
addressing the physical issues. There could be problems getting to and from the
bus stop and sometimes there were problems with ramps and small wheelchair
spaces. However, it was generally the case that it was physically possible to
complete a journey without excessive difficulties.
Perhaps the most striking issue to emerge from the research was the role that
psycho-social factors played in affecting mobility-challenged people’s quality of
experience of using public buses, in particular, the impact of the attitudes and
behaviour of the driver and of other passengers.
Bad experiences of this nature were the most frequently cited reasons for not
enjoying a bus journey or for not using the bus at all. Previously, the emphasis of
accessibility research and improvements schemes has been on the physical
elements of accessibility. While these are certainly extremely important, the
outcomes of our research suggest that psychosocial issues are equally, perhaps
20 Nickpour et al.
even more, so. This observation mirrors those within the field of design generally
where there has been increasing attention in recent years on psychosocial issues
and their emotional consequences (Norman, 2005).
2.4.2 Special Service Versus Public Service

As part of this research we also looked at people’s experiences with door to door
transportation schemes for mobility challenged people within London. These
included Dial-a-Ride, a minibus-based service which picks up passengers at their
home and takes them to a pre-requested destination. This service was very popular
with users. In particular they enjoyed the friendly atmosphere on the minibus and
the friendly, attentive and considerate behaviour of the driver.
Mobility-challenged users praised the drivers for their empathy and
understanding, for their cheerfulness and for making them feel valued and
welcome whenever they used the service. They mentioned how much they looked
forward to the social aspects of using the service and for the enjoyable
conversations with other passengers. A challenge is to try and recreate some of
these benefits on public buses and to put into place approaches and schemes that
will help to foster a positive ambience.
2.4.3 Negative Interactions

It should be emphasised that the picture is not entirely negative; Field research
supported the fact that many of the drivers have an excellent approach to
interacting with mobility-challenged people. They are friendly, welcoming,
informative and help make the journey a great experience. Similarly, many
teenagers are polite, well-behaved and kind towards other passengers. However,
this was mainly the result of each individual’s intrinsic motivation and personal
codes of conduct.
Nevertheless, it is also important to recognise that there are genuine problems
with some bus drivers’ and teenagers’ attitudes and behaviours. Negative drivers’
attitudes were observed and reported, such as being rude and uncommunicative
towards mobility challenged people. Also, in some cases, some teenagers’
behaviour appeared inconsiderate and liable to make people feel uncomfortable.
The effects of this negative behaviour tend to extend beyond the specific
incidents that occur. When service users encounter a bad experience, they will
remember this and will have a doubt in their minds about the quality of their
experience next time.
This uncertainty can have a very powerful and negative effect. Even if people
subsequently have positive experiences, the memory of the previous bad
experience can create a sense of doubt - will this happen again? This doubt can
make people question whether they want to use the bus again and leave them with
some negative feeling for the duration of their travel. Moving forward, the
challenge is to find effective ways of improving the ambience on board and

tackling some of the psychosocial issues that have been identified.
2.5 Conclusions and Recommendations

There is need for a ‘Mentality Shift’ when addressing accessibility in public
transport. This study suggests and highlights ‘psycho-social’ inclusion as the key
area of focus. The findings suggest accessibility and inclusivity issues affecting
public bus services fall into three broad categories: Physical, Psycho-social and
Operational.
Physical issues are to do with the design of the bus and the built environment
and are the ‘typical’ issues considered when looking at accessibility. Findings
suggest the key physical barriers identified include Getting to bus-stop, Space
availability and priority on bus and Ramp technology & reliability.
Psycho-Social issues are the ‘soft’ issues associated with the quality of people’s
travel experience. Findings suggest the key psycho-social barriers identified are
Ambience, Awareness and empathy and Communication.
Operational issues concern the running of the service and cross-organisational
strategies and regulations. The key identified operational barriers are Key
Performance Indicators. Public bus service KPIs currently appear to focus only on
efficiency rather than quality, inclusivity and pleasurability of service.
The results indicate that it is the psycho-social issues that seem to be proving
the biggest barrier to using public buses, in particular for mobility-challenged
people. Addressing these issues requires a focus on people. It involves making
them aware of the effect that their behaviour is having, convincing them to change
it and giving them the skills and insights needed to do so. It also involves creating
a desirable ambience throughout the bus journey, making the public transport
experience not only efficient but also pleasurable.
Overall - including both physical and psychosocial factors - the following nine
recommendations are proposed as key principles for improving mobility
challenged passengers’ experience of public bus travel.
Create an inviting and friendly experience of the bus service. Perceptions about
bus travel influence people’s decisions about whether to take the bus and the
emotions associated with anticipating using it. Mobility challenged people should
be confident that their bus journey will be a positive experience.
Make bus stops reachable. Getting to and from the bus stop is, generally, the
biggest physical barrier to bus travel for mobility challenged people. Making bus
stops more reachable would significantly increase the numbers of people who
could access public buses.
Make all bus stops fully accessible. Once at the stop, mobility challenged
people should be accurately informed about when the bus will arrive. The design of
the stop should also facilitate quick and easy ingress for them.
Promote and facilitate positive behaviour amongst passengers. Interactions
with other passengers should be positive and friendly throughout the bus journey.
22 Nickpour et al.
Ensure that key aspects of the bus are fully operational. The aspects of the bus
that affect accessibility should be fully operational at all times. Mobility challenged
people should be confident that their journey will run smoothly and efficiently.
Ensure that all users have a safe and comfortable space. All mobility
challenged users should have a safe and comfortable space in which to complete
their journey. They should be able to move into and out of this space easily.
Welcome mobility challenged people aboard. Drivers should warmly welcome
mobility challenged people aboard the bus. They should communicate clearly and
cheerfully with them throughout the journey.
Set off and drive smoothly. Ensure that mobility challenged people are settled
before moving off. Make sure that this is done smoothly and that the drive is
smooth and controlled throughout the journey.
Provide information clearly through multiple channels throughout the journey.
Mobility challenged people should be clear about when the bus is approaching
their stop and have plenty of time to prepare to exit.
This research project was commissioned by London Borough of Hillingdon and
Transport for London. The authors would like to thank all local participants in the
project and the user research team including Murtaza Abidi, Penelope Bamford,
Thomas Wade and Jennifer McCormack.
2.7 References
Dray SM (1997) Structured observation: Practical methods for understanding users and their
work in context. In: Proceedings of CHI 97 Workshop on Human Factors in Computer
Systems, SIGCHI, Atlanta, GA, US
GLA (2010) Accessibility of the transport network. Greater London Authority, London, UK
London TravelWatch (2009) TfL performance report. London TravelWatch, London, UK
London TravelWatch (2010) Bus passengers’ priorities for improvement in London.
London TravelWatch, London, UK
Moore P, Conn CP (1985) Disguised: A true story. Word Books, Waco, TX, US
Nickpour F, Jordan PW (2011) Inclusive bus travel in Hillingdon: Assessing accessibility.
Technical Report, Brunel University, Brunel, UK
Norman DA (2004) Emotional design: Why we love (or hate) everyday things. Basic Books,
Basic Books, NY, US
Transport for London (2011) TfL accessibility guide. Available at:
http://www.tfl.gov.uk/gettingaround/transportaccessibility/1171.aspx (Accessed 13
August 2011)
Chapter 3
Safe and Inclusive Design of Equipment

Used in the Minerals Industry
T. Horberry and T. Cooke
3.1 Introduction
The focus of this paper is upon the application of both safe and inclusive design to
equipment used by operational and maintenance personnel in mining. It begins by
introducing the minerals industry and outlining two important human-related issues
that will greatly impact upon the design of future mining equipment. The paper
then focuses on the importance of safe and inclusive design in this domain, and
outlines a task-orientated risk assessment and design process called ‘OMAT’ that
was developed by the authors. A series of semi-structured interviews with
designers of mining equipment are then presented: they focus in particular on how
designers currently obtain user-centred input and how the OMAT process might be
integrated into their design practices. Finally, conclusions regarding the future safe
and inclusive design of equipment (especially automated equipment) in this
domain and why user-centred design processes should be of central importance to
mining are presented.
3.1.1 An Overview of Mining and the Minerals Industry

The minerals industry is a significant worldwide employer. For example, in
Australia it employs approximately 136,000 personnel. Mining occurs across
virtually the whole globe, with major areas in South Africa, North and South
America, Australia, China and much of Europe. The worldwide injury, ill-health
and fatality rates vary greatly, ranging from usually single figure deaths per annum
in Australia through to many hundreds being killed in many third-world countries
(Simpson et al., 2009). Whatever the precise figures, the minerals industry is a
major global employer with many high hazards that can cause injuries and fatalities
unless well managed (Komljenovic and Kecojevic, 2007).

24 Horberry and Cooke
Looking in more depth at the elements of the minerals industry system, there is
no ‘typical’ arrangement that is used at the majority of mine sites (Sanders and
Peay, 1988). Instead, it is a complex sociotechnical system where people,
procedures, environments and equipment need to interact safely and efficiently.
The main elements here include (Horberry et al., 2010):
• an increasingly diverse group of people employed;
• a wide assortment of different jobs, tasks and roles;
• many different equipment manufacturers and suppliers;
• different worldwide mining companies;
• a wide array of national laws, regulations, and guidelines;
• different procedures, rules, practices and cultures at individual mine sites;
• differences in the built environment and precise mining method used;
• uncertainties in the natural environment.
3.1.2 Mining Equipment

Focusing on the equipment element, there is a discernable trend in most advanced
economies for mining equipment parameters to be improved in that the equipment
needs to be bigger, stronger, quicker, safer and more reliable (Horberry et al.,
2010). The images in Figure 3.1 show the size and complexity of two examples of
mobile mining equipment.
Figure 3.1. Examples of mobile mining equipment
In tandem with this, there are continual equipment-related operational

challenges, including the ever-present balance between safety and production. It is
not within the scope of this paper to describe the range of equipment employed, but
it certainly is within the scope to briefly describe two general human element
challenges that impinge upon designing and deploying safe, inclusive and fit-for-
purpose equipment. These are: the ageing workforce and the increased
development and deployment of automation/new technologies in mining.
Safe and Inclusive Design of Equipment Used in the Minerals Industry 25
3.1.2.1 An Ageing Workforce

Over twenty years ago it was noted that the average age of the mining workforce
was getting younger (Sanders and Peay, 1988). In most industrialised countries this
situation has now reversed: the minerals industry workforce is on average now
getting older, and fatter. Of course, this mirrors many general trends in Western
society; however, it does present a few specific issues for the design of mining
equipment, especially to make such equipment inclusive and safe. Given the nature
of the equipment in use then extreme ageing is not an issue in mining, but there are
still many issues for safe and inclusive design that result from an older workforce.
These include (adapted from Horberry et al., 2010):
• Increased difficulty in learning new skills. Older workers do not automatise
tasks as easily as their younger counterparts. This has implications for
mining technology and automation use (e.g. remote control of a rock
breaker) where skill requirements may change over time and require new
automatic, over-learnt operating procedures.
• Reaction time increases, especially in reacting to unexpected stimuli. This
is a particular concern for mobile mining equipment (e.g. haul trucks)
operating in the complex transport environment in most mine sites.
• Loss of muscular strength, endurance and tone. Although many manual
tasks in mining have been eliminated, they are still present in many
equipment maintenance tasks. This problem is exacerbated if significant
weight gain occurs in these older maintenance workers.
• Visual function changes including loss of precision, difficulty in focusing
on near objects and declines in visual acuity and contrast sensitivity.
Given the automated, round-the-clock nature of mining then this can
present particular problems in bad lighting conditions or with a complex
background.
3.1.2.2 Increasing Development and Deployment of New Mining

Technologies and Automation
Mining is already highly mechanised, but it seems certain that there will be more
remote control and/or automation of mining equipment in the future. Indeed,
automation in mining is now finally permeating into many cutting-edge sites.
Although automation progress began initially during the 1960s, today many mining
companies are investing heavily in automation initiatives (Lynas and Horberry,
2011).
As in other industries, this may change the types of human element inputs
required (Sheridan, 2002). For example, less manual operational tasks (at least when
the equipment is working correctly) and many operators becoming more of a passive
monitor of the system, rather than an active controller or driver of it.
More broadly, the increasing uptake of mining automation presents many
opportunities, but also many challenges related to safe and inclusive design and
effective human systems integration (Lynas and Horberry, 2011). The role of the
human in the overall mining system may change, but is still a central part rather than
an optional extra. Thus, developing operator-centred approaches for the design and
integration of new/automated mining technologies is a key priority area for the

technology to be successful: central to this is safe and inclusive design.
3.1.3 The Importance of Safe and Inclusive Design of

Mining Equipment
The design of mining equipment plays a critical part in the safety and efficiency of
tasks that are conducted by operators using it. However, as much as in virtually any
other occupational domain, the design of mining equipment is still heavily
technology-centred rather than user-centred (Horberry et al., 2010). This has been for
a variety of reasons in mining, including:
• Designers sometimes think they can use their knowledge/common
sense/intuition, or they rely purely on designing to standards. This is
particular acute in the minerals industry, where mine site access for designers
can be a significant limitation (EMESRT, 2011).
• Adding a user-centred focus may be thought to alter an agreed design
process. Similarly, an older version of a system may already be in place, and
piecemeal alterations are subsequently made (for example, the design of haul
trucks).
• The benefits and costs of using a user-centred approach for equipment design
are not clear (Burgess-Limerick et al., 2007).
Because some mine equipment designers (and their customers) still view human
element constraints to be less significant than technical challenges (such as
equipment payload) there is a tendency not to adequately consider human factors in
the equipment design life-cycle process, and it is common to see human factors
concerns being passed from one phase to the next (Horberry et al., 2010). For
example, during conceptual design, if the analysis does not adequately capture user
requirements then subsequent inadequacies are hard to resolve in the detailed design
phase. Human element problems (such as usability) that remain after the system has
been designed cannot therefore be easily resolved during equipment build or
implementation phases. This is a particular problem for mining equipment that still
requires considerable human intervention in its maintenance and operation (e.g.
mobile mining equipment maintenance such as changing air filters).
Indeed, where such user-centred design does exist, it focuses more on safe rather
than inclusive design. Safe design (sometimes known as safety in design, or
prevention through design) has begun to receive an increasing amount of attention in
both the scientific and occupational safety domains (including in mining, albeit
slowly) and is generally applied to products and equipment. As the name implies, it
involves occupational health and safety (OHS) through the original design, not safety
by procedure or through retrofit trial and error.
“Safe Design is a design process that eliminates OHS hazards, or minimises potential
OHS risk, by involving decision makers and considering the life cycle of the designed-
product.”
(Safe Work Australia, 2011)

In contrast, inclusive design is still in its infancy in mining. As an occupational,
high-hazard domain, it might be argued that a ‘conventional’ inclusive design
process is in need of modification as the focus here is not on mainstream
commercial products. However, the earlier-described changing workforce
demographics (e.g. older operators) and the design possibilities raised by the
uptake of automation should increase the focus on designing for an increasingly
diverse user group that will engage in a wide range of new tasks.
3.1.4 The Operability and Maintainability Analysis

Technique (OMAT)
To help ensure safe and inclusive design, the authors of this paper recently
developed and evaluated a process that could be used by mining Original
Equipment Manufacturers (OEMs) (Cooke and Horberry, 2011). More specifically,
OMAT is a task-oriented risk assessment process that focuses on human factors
risks related to mobile mining equipment design.
The starting point of the work was that poor equipment operability or
maintainability can produce major safety and performance disbenefits. As such,
many mining incidents and accidents are due to equipment design inadequacies,
either in maintainability or operability, and are therefore theoretically preventable
(Horberry et al., 2009). The behaviour of the operators and maintainers is largely
shaped by their tasks which are, in turn, partly shaped by the equipment’s design.
Therefore, in order to create the safest equipment possible designers must predict
how their designs will shape the behaviours of miners in different sites around the
world.
With this in mind, a number of mining companies started discussing an
approach to improve the human factors design of mobile mining equipment at an
OEM level. This eventually resulted in a multi-company industry initiative known
as EMESRT; the purpose was to establish a process of engagement between OEMs
and mining customers to promote the development and adoption of leading
practice user-centred designs (EMESRT, 2011). The OMAT tool was developed by
support from EMESRT (Horberry et al., 2009).
The OMAT process has been developed to align with existing OEM design
processes (including major design milestones). OMAT has also been developed
specifically to provide a user-engagement processes to identify and assess the risks
in the design of mining equipment. Investigating such risks in operational or
maintenance tasks involves six OMAT steps (after an initial step zero, to define the
scope of the work) that are heavily dependent on OEM and mine site user
engagement. A summary of these six stages is presented below.
1. Based on a comprehensive list of all operations and maintenance tasks

performed using the equipment, the critical tasks are prioritised.
2. The constituent steps in these priority tasks are described and analysed. In
effect this is a task analysis, whereby each task step and its order is
uncovered, including deviations, short cuts or different methods of task
completion for different user-groups.
3. Risks at each of the task steps are identified. Using the types of matrices
commonly used in the mining industry the risks are identified, noting any
current controls employed (e.g. guard rails for working at height).
4. Solutions are developed for the risks identified in stage 3. These should be
primarily design solutions that eliminate risk.
5. Feedback received. The solutions developed in stage 4 are further
developed by the OEMs. However, to continue the process of user-centred
design, these solutions are then evaluated by mine site users.
6. A risk register is maintained to keep track of the whole process. This is of
particular value for future design iterations of the equipment.
The development and largely positive evaluation of the OMAT process to help
promote safe and inclusive design of existing mobile mining equipment has
resulted in important design changes being made (see Cooke and Horberry, 2010).
But a comparatively neglected area to date has been an assessment of designers’
opinions of the process, and how it can be integrated with their existing safe and
inclusive design processes.
3.2 Interviews with Mobile Mining Equipment

Designers
3.2.1 Aim
The aim of the study was to gauge the opinions and work practices of mobile
equipment designers regarding user-centred design processes. In particular, it
compared their current design methods to the OMAT task-based model and
explored the designers’ current methods of gaining end-user feedback (i.e. from
mine site operational and maintenance personnel).
3.2.2 Method
A series of semi-structured interviews were held with personnel employed by a
major manufacturer of mining equipment. Fourteen people who worked for the
manufacturer were interviewed; all interviews took place in Finland. The
participants worked in a variety of design-related roles including design engineers,
project managers, safety experts and user interface specialists.
Rather than asking a rigid set of pre-defined questions, the interviews were
topic based, allowing the interviewer to prompt, clarify and further probe the
thoughts and practices of those being interviewed. Three topic areas were covered:
Topic 1: The current design situation. Interviewees asked about the formal and
informal human-related methods used to create and assess safety of the equipment.
Specifically, they were asked to consider methods that considered end-user
interaction. For each method noted, the interviewees were asked to describe the
strengths and weaknesses of that technique.
Topic 2: OMAT review. The OMAT process was then described to the participants
(many already had a basic knowledge of OMAT). The participants were asked to
judge and predict the strengths and weaknesses of this technique, and to state how
it might be integrated within their overall design processes.
Topic 3: End-user input. Of course, gaining end-user input is a common theme in
safe, participatory and inclusive design. However, this has increasingly become
difficult as the design, manufacture and use of equipment commonly occurs in
different countries across varying cultural and language barriers. As an example of
this, the interviews were conducted in Finland for equipment that would be used
ultimately in Australia or elsewhere.
This topic explored the currently-used methods of gaining end-user input into
their design processes, especially as there was no homogeneous end-user group. It
then explored two practices that could be used within the overall OMAT
methodology to potentially gain more comprehensive end-user feedback: the
inclusion of more (and varied) end-users in the OMAT workshops and the
collection of video footage of tasks being performed with current equipment.
3.2.3 Results
The results are described in terms of the three topics mentioned above. For reasons
of space, only summary findings are reported here.
Topic 1: The current design situation. The interviewees reported a large number of
broadly human-centred methods to improve the safety of their equipment. They
also reported that considering the end-user was a constant informal consideration.
However, only four methods were noted that specifically included some
consideration of the end-user.
Two of these methods were only used by specific design specialists. The first
was an “Ergonomic Checklist for the Operator’s Cabin”. This primarily involved
ensuring that the adjustability of elements like reach distances to controls and chair
design for weight of operator was accounted for using anthropometric data. The
second method was in-field usability observation. However, this was limited to
new digital controls used during operation (primarily of automated equipment).
The interviewees thought that these methods were potentially useful; however, the
scope and application of them was extremely limited.
The third method involved direct customer feedback from the use of current
products. This was noted to be beneficial as many issues only emerged during
equipment use. However, the major issue related to the ‘voice of the customer’
was that the input was only text-based and needed to pass through many hands
before it reached the appropriate designer. By this stage it was commonly difficult
to understand the precise issue and the person who raised it was not contactable.
Therefore, only easily explainable and understandable issues fed through the
current system with consistent success.
The fourth, and most noted, method was that of a risk assessment workshop
using qualitative risk matrices. The issues primarily came from equipment
standards and were a mix of hazards and requirements. The primary positive issue
noted about this method was that the workshop forced conversation amongst the
stakeholders. The most significant negative issue noted was that the discussion
focused only on the rather restrictive set of issues mentioned in existing standards.
This commonly did not involve the end-user but, rather, ‘have we met the
standard?’ This meant that significant issues related to operator interaction could
be missed, and innovation might be stifled. Furthermore, the interviewees noted
that extremely rarely did the method actually lead to design changes. As such, the
interviewees mostly viewed this method as a legal cover rather than a useful design
method.
Topic 2: OMAT review. Despite the potential bias that might have been present
(where one of the originators of the process was conducting the interviews), the
interviewees’ opinions of OMAT were largely positive. In particular:
• they welcomed the task-based, user-centred approach;
• they found the whole OMAT process to be easy to understand;
• it was viewed as comprehensive;
• they predicted that it might encourage innovation in design solutions.
However, some of the more negative features noted were:
• the OMAT process might take a long time (especially in workshops);
• it might be difficult to maintain/update this method;
• some concerned was expressed that it does not link sufficiently to
standards. Although this was not fully articulated, there was more of a
general feeling of unease about how much it became the workshop’s
judgement (rather than an ‘objective’ standard) and having that judgement
documented for a lawyer to subsequently investigate.
Topic 3: End-user input. The inclusion of a variety of end-users in design
workshops (whether using OMAT or other methods) was reasonably positively
received. This was because operational knowledge was seen to offer valuable
design opportunities by knowing ‘what really happens’. However, there were also
significant negative issues noted (that are also applicable to many other
participatory, inclusive or safe design processes):
• the end-user inputs received are limited to the knowledge of that person;
• legal issues (‘what happens if we don’t take their advice?’);
• disclosure of information about design, so market value might be

potentially reduced;
• they already have in-house operational ‘experts’ involved so current end-
users might not add much extra information;
• pragmatically, and certainly when considering universal design, the
potential users come from across the world and speak different languages.
This was perceived to be a potentially costly exercise for little gain.
The technique of obtaining end-user inputs by means of video recording of
current tasks was viewed by every interviewee as an extremely beneficial addition.
In particular, it was opined that video records provided objective visual
information that is easily shared, showed the designers how their equipment was
actually used, acted as a memory aid for those designers who had actually been to a
mine site and more generally could be an excellent reminder that real people will
eventually have to use the equipment.
Video records are, of course, not as an adequate replacement of comprehensive
end-user feedback; the interviewees noted that they were unable to ask the person
about the task, so had to interpret the actions on the video. Equally, other design
options could not be explored, and if a design is changed significantly then the task
itself changes. Finally, video records were only useful in the latter stages of the
design process, so a designer would have to wait for the equipment to be out in the
field before they can get this first iteration of the feedback.
3.3 Conclusions
Including end-user input in some form in the design of equipment is a laudable
goal in virtually any domain. As indicated in other domains that have successfully
integrated automation (e.g. aviation), unless new technology in mining takes into
account the human element that will ultimately operate or maintain the systems,
then it is unlikely that such technology will flourish (c.f. Sheridan, 2002).
This paper has hopefully shown some of the challenges and opportunities for
those mining equipment designers who have a human-centred focus. Some of the
challenges include designers actually getting access to mine sites, and obtaining
appropriate end-user input, due to the widely diverse workforce involved. Not
surprisingly, a disjointed group of user-centred methods currently exists, and these
are employed to varying degrees by mining equipment designers. Despite this, safe
design in some guise is certain to become more firmly entrenched in the design
practices of larger OEMs. Consequently, it is anticipated that the task-based nature
of OMAT will result in it being more widely accepted and deployed, especially for
high frequency/high hazard tasks. A criticism of OMAT was that the method could
be quite time-intensive, especially when involving extensive end-user workshops
that needed to be formally documented. One way to address such a criticism would
be to employ a streamlined version of the method that is still task-based and
participatory, but only focuses on design issues of the highest priority tasks in a
less formally documented workshop-style setting. Also, obtaining end-user inputs
by means of video recording of current tasks with the equipment is another way of
streamlining the OMAT method whilst still retaining its task-based approach.
Inclusive design of mining equipment is far less well-developed and well-
accepted than safe design. Indeed, philosophically, it might be maintained that
specialised and hazardous mining equipment should not be designed for universal
use unless reasonable controls (including training and following site procedures)
are deployed. Still, designing accessible and usable mining equipment without
excessive adaptations for the worldwide minerals industry user-group is certainly
of importance for both safety and productivity. It is anticipated that user-centred
design tools such as OMAT can be expanded and further employed to help achieve
this goal.
3.4 References
Burgess-Limerick R, Straker L, Pollock C, Dennis G, Leveritt S, Johnson S (2007)
Participative ergonomics for manual tasks in coal mining. International Journal of
Industrial Ergonomics, 37: 145-155
Cooke T, Horberry T (2011) The operability and maintainability analysis technique:
Integrating task and risk analysis in the safe design of industrial equipment. In:
Proceedings of the International Conference on Ergonomics and Human Factors 2011,
Stoke Rochford, UK
EMESRT (2011) Earth moving equipment safety round table. Available at:
http://www.mirmgate.com/index.php?gate=emesrt (Accessed 8 August 2011)
Horberry T, Burgess-Limerick R, Steiner L (2010) Human factors for the design, operation
and maintenance of mining equipment. CRC Press, Boca Raton, FL, US
Horberry T, Sarno S, Cooke T, Joy J (2009) Development of the operability and
maintainability analysis technique for use with large surface haul trucks. Australian Coal
Association Research Program report. Available at: http://www.acarp.com.au/
abstracts.aspx?repId=C17033 (Accessed 8 August 2011)
Komljenovic D, Kecojevic V (2007) Risk management programme for occupational safety
and health in surface mining operations. International Journal of Risk Assessment and
Management, 7(5): 620-638
Lynas D, Horberry T (2011) Human factors issues with automated mining equipment.
Ergonomics Open, 4(Suppl 2-M3): 74-80
Safe Work Australia (2011) What is safe design? Available at:
http://www.safeworkaustralia.gov.au/SafetyInYourWorkplace/SafeDesign/Understanding
/Pages/WhatIs.aspx (Accessed 8 August 2011)
Sanders MS, Peay JM (1988) Human factors in mining (IC 9182). Department of the
Interior, Bureau of Mines, Pittsburgh, PA, US
Sheridan T (2002) Humans and automation. John Wiley, New York, US
Simpson G, Horberry T, Joy J (2009) Understanding human error in mine safety. Ashgate
Press, Farnham, UK
Chapter 4
Deploying a Two-player System for Arm

Rehabilitation in Schools
R.J. Holt, A.P.H. Weightman, J.F. Gallagher,
N. Preston, M.C. Levesley, M. Mon-Williams and
B. Bhakta
4.1 Introduction
Cerebral Palsy (CP) is the commonest cause of disability among children in Europe
(Johnson, 2002). Its effects and severity can be extremely varied, but a
combination of arm impairments through weakness or spasticity and sensory
deficits is common and can (i) significantly impair the ability of individuals with
CP to carry out daily activities and (ii) create significant social barriers (Imms,
2008). Therapy is often used to aid the acquisition of motor skills, particularly in
childhood, but a lack of physiotherapy resources means that this is often delivered
through a self-managed home exercise programme. Exercises are frequently dull
and repetitive, and children often lack the motivation to carry out these exercises,
leading to poor compliance with the prescribed plan (Chappell and Williams,
2002).
One solution to this is the use of Interactive Computer-Play (ICP)-based
therapy (Sandlund et al., 2009), where therapy is delivered as a game through a
computer-interface. Prior research at the University of Leeds has led to the
development of a game-based system for home rehabilitation of upper limb
impairment (Weightman et al., 2011). However, feedback from the participants in
that project indicated that they preferred to play games with friends and suggested
that this would improve their motivation further. Accordingly, a multiplayer ICP-
based therapy system for upper limb rehabilitation intended for use in schools has
been developed, and this paper describes both the system and the results of its
initial deployment in a school environment. A brief overview of the background to
ICP-based therapy and arm rehabilitation is given, followed by a description of the
system itself, and finally the outcomes of deploying the system in a school
environment for an eight-week period, including the amount of usage, feedback on
the games used, and user comments received.

34 Holt et al.
4.2 Background
ICP-based therapy has been growing in popularity in recent years, particularly with
the advent of home computers and videogame consoles and the growing popularity
of videogaming as a pastime in the last few decades. The development of consoles
which use movement-based interaction in videogames, most notably the Nintendo
WiiTM, has led to great interest in their use as a means of encouraging physical
activity among children and making rehabilitation enjoyable (Deutsch et al., 2008;
Lanningham-Foster et al., 2009). The use of off-the-shelf videogame consoles in
rehabilitation has many benefits, as they enjoy the economies of scale of mass
production, do not require specialist development, and games are already designed
first and foremost to be enjoyable. However, they also have limitations: they are
not necessarily accessible to players with more significant arm impairments; they
do not provide the assistive force that a physiotherapist would provide (which is
important in extending a patient’s capabilities); the games are not necessarily
designed to provide therapeutic benefits; and these systems do not enforce
compliance with a therapeutically desirable trajectory, which means that players
can get away with making motions that are successful in the game, but do not
provide therapeutic benefit.
Research at the University of Leeds has promoted the use of Assisted
Movement Devices as an adjunct to therapy, whereby a robotic system is used to
provide both the assistive force and to encourage compliance with therapeutically
desirable trajectories. Previous research has developed systems that use this
approach with stroke patients in a clinical environment (Jackson et al, 2007) and
for children with in a home environment (Weightman et al, 2011). In response to
feedback gathered from the latter project, a two-player system has been developed,
with the aim of deploying it in a school environment, where it would be easier to
find multiple players. Social interaction such as co-operation and competition in
games has long been identified as a motivator for playing (Malone and Lepper,
1987), and continues to be recognised as an important aspect of making games
enjoyable (Sweetser and Wyeth, 2005). However, it does raise significant
challenges in ICP-based therapy, as different players will have different levels of
impairment (and in some cases, none at all), making it difficult to create a level
playing field. The next section reviews the system developed, and how it addresses
some of these issues.
4.3 The System

The original home-based system had been based around a Microsoft SideWinderTM
joystick, which had been adapted so that its force feedback system would provide
the assistive force required to help players make the required movements
(Weightman et al., 2011). This plugged via USB into a conventional PC, where the
specially designed games would run. It was originally envisaged that the new
system would take a similar form, with perhaps four to six such joysticks plugging
Deploying a Multiplayer System for Arm Rehabilitation in Schools 35
into a PC to allow multiplayer gaming. However, a number of factors meant that

this approach was not feasible.
Firstly, the Sidewinder was not able to provide as much force as desired,
meaning a move towards a new design with larger motors and bespoke control
software developed in LabVIEW and delivered via a National InstrumentsTM cRIO
(Compact Reconfigurable Input-Output) controller. Furthermore, in working
closely with teachers, children and parents to develop the system, it soon became
apparent that class time was at an absolute premium for teachers, and they argued
that they could not afford to spend take even a few minutes out of class time setting
the system up and getting started without the children losing interest. Space was
also at a premium, meaning that a four to six player system would be prohibitively
large - teachers felt that this was redundant as they rarely had more than one or two
children in a class requiring physiotherapy and were reluctant to take many
unimpaired children out of lessons to participate in the games.
Accordingly, the system was designed as a self-contained unit with two
joysticks, two monitors, a PC and cRIO as shown in Figure 4.1. The system was
wheeled so that it could be easily moved between classrooms or out of the way as
needed, and required only a single button press to start up or turn off, booting
directly into the games as needed.
Figure 4.1. The system
To address the need for a level playing field between different players, an
Adaptation to Player Performance Algorithm (APPA) was developed. Four games
were developed, representing different combinations of competitive and
collaborative, sequential and simultaneous play. They all centred around the same
premise, delivered through simple cut scenes: the players are monkeys trying to
rescue their friends from a hungry crocodile in order to provide elements of fantasy
36 Holt et al.
and curiosity as recommended by Malone and Lepper (1987) and Sweetser and
Wyeth (2005); and all were based around the simple back-and-forth movement of a
conventional reach/retrieve exercise. The APPA first involved the players carrying
out a simple single-player assessment task at the start of each session, in which
each player guided their ‘monkey’ around the screen and tried to collect as many
‘bananas’ as possible in the time available. The system then adjusted the amount of
assistance provided based on their performance in this task. The four actual games
were developed with a user group of children with cerebral palsy who had
participated in the previous home-based project (Weightman et al., 2011). These
children gave feedback and made comments on early iterations, evaluating the
initial concepts and gameplay proposals, giving feedback on early prototypes and
testing the near final games for usability. The final four games selected were:
1. Van Game (Simultaneous, Cooperative): Players work together to collect
bananas and destroy the crocodile’s van before s/he escapes with their
monkey friends.
2. River Game (Simultaneous, Competitive): Players race against each other
to collect the most bananas as they race along a winding river.
3. Chase Game (Sequential, Competitive): Players take it in turns to find
their way through a maze, one playing as the crocodile, the other as a
monkey - the player with the fastest time wins.
4. Maze Game (Sequential, Cooperative): Players work together to find a way
through a maze before time runs out, as each player collects bananas so
new paths open to the other player.
Screenshots from these games are shown in Figure 4.2. Each game also had a
single player variant, in which the player raced against the clock, rather than the
other player.
4.4 Outcomes of Deployment

To evaluate the feasibility of using the system in a real school environment, the
system has been field-tested in seven schools with an overall total of eight children
with CP aged between 8 and 12 years (to date). At this stage, the purpose of this
testing is to assess whether the system can actually be delivered and used in a
school without supervision by the research team, rather than assessing whether the
system delivers therapeutic benefits.
4.4.1 Process
A total of four systems were built. The systems were deployed into schools for two
periods of four weeks, with a one-week “washout” period in between. In one
period, the system was used in a single-player mode; in the other, it was used in a
multiplayer mode. Half the schools were randomly assigned to use the single
player mode first, while the other half used the system in multiplayer mode first.
As there were seven schools and just four systems, the systems were deployed in
four schools first (Schools A, B, C and D) and in the remaining schools thereafter.
a) Van Game b) River Game
c) Chase Game d) Maze Game
Figure 4.2. Games for use with the system
The research team delivered the system to each participating school, explained
to the member of staff who would be given responsibility for the system how to set
it up and play the games, and then left the system under the supervision of the
school for the four-week period. Teachers were introduced to the goals of the
research project, made aware that the system was intended as a way of delivering
therapy through gameplay and informed that the version of the system delivered
was a prototype for evaluation with the aim of assessing the feasibility of its use in
a school environment. The initial explanation was supplemented by an instruction
manual, and a contact number to call in case of significant problems. The aim was
to see whether and how each school would use the system in practice, without the
research team’s input, so they were permitted to make whatever use of the system
they saw fit.
The systems gathered details of the amount that each child played, and a
feedback questionnaire based on Read’s (2008) Smileyometer was used for each
child to rate each of the games and the assessment task. Finally, a debriefing
38 Holt et al.
questionnaire was used to capture the views of adults and children about the
system, any problems or benefits they encountered, and what might be done to
improve it in the future. As a result of difficulties in getting time in teachers’
diaries for interviews in the early phase of the project, it had been agreed that the
most effective way of gathering information was to provide paper feedback
questionnaires with a stamped self-addressed envelope that the school could return
to the researchers at their convenience, rather than attempting to arrange formal
feedback interviews.
4.4.2 Usage
Table 4.1 records the amount of usage made of the system for each child at each
school (note that School C had two children), in terms of the number of days upon
which the system was used, and the mean length of play on those days. In addition,
it shows the mean number of days on which the system was used across all the
children, and the mean length of all their sessions. Data were only recorded for
children with cerebral palsy as they were the target users of the system: the project
was not concerned with the amount of therapy delivered to children who did not
require it.
Table 4.1. Usage of system by school and child
School Child Days Used Mean Session Days Used Mean Session
Single Length Single Multi- Length Multi-
Player (of Player (mins) player (of player (mins)
possible 20) possible 20)
A 1 10 6.68 19 9.38
B 2 5 5.26 13 10.2
C 3 15 21.5 19 17.2
C 4 15 17.7 16 19.0
D 5 12 8.71 10 9.51
E 6 15 3.38 12 6.20
F 7 13 15.5 10 9.70
G 8 10 13.3 16 20.5
Overall Mean 11.9 12.7 14.4 13.4
All the children made some use of the system and with just one exception
(Child 2’s single player phase), it was used on at least half the available days. In
some phases, days were lost owing to school closures in bad weather and bank
holidays, but these were accepted as part of a realistic snapshot of how much use
the system received in practice (as these were all factors that might limit its use
they are not corrected for in the table above). On average, multiplayer mode was
played more often and for longer than the single player mode, despite the difficulty
in having to take children without arm impairments out of lessons to accommodate
this - though with such a small sample, the difference was not statistically
significant.
4.4.3 Game Ratings

Getting schools to complete and return questionnaires proved extremely difficult,
and several rounds of chasing up were required. A total of fourteen smileyometers
were returned, though we know that a total of thirty-two children across the seven
schools (including the eight with CP) used the system altogether. Unfortunately,
the questionnaires were not always correctly labelled, making it impossible at some
schools to determine which feedback was from children with CP and which from
children without. Accordingly, the results presented here represent the overall
feedback from all the children that provided Smileyometers rather than trying to
distinguish between the different groups of users. The results for each game,
including the assessment task, are shown in Figure 4.3.
Figure 4.3. Smileyometer ratings for the four games and assessment tasks
40 Holt et al.
Overall, the two co-operative games - the Van and Maze Games - proved most
consistently popular, with the Maze Game receiving no negative ratings (Not Very
Good or Awful), and the Van Game receiving just one not very good. The River
Game caused the greatest split in opinion, receiving the most Brilliants, but also
receiving the most negative ratings of the four games. The Chase Game also
received mixed feedback, but with a more even split across the ratings. The
Assessment Task also split opinion, receiving more negative ratings than any of the
games, but also surprisingly receiving a large number of positive ratings, with only
the River Game receiving more positive ratings. It is worth noting that the ordering
of the games varied from child to child quite significantly as well: each game was
most popular for at least two children, and least popular for at least two others. On
the whole, this suggests that the games were well-received, although there were
substantial variations in individual preferences, and several children did suggest
that more games would be welcome to provide greater variety.
4.4.4 Qualitative Comments

Feedback comments from the school staff responsible for the system focused
consistently on the size of the system, and ease of setup. One school indicated that
manoeuvrability was not important, as the size of the system meant that it could
not be moved anywhere else. The other schools indicated that being able to move
the system around easily was very important, as it needed to be used in different
classrooms, or moved out of the way when not in use. Every school was able to
find space for the system, and none felt it was too big, but all agreed that a system
any larger than this would be untenable.
Ease of setup was also an issue, with all schools praising the ease with which
the system could be started, but also reporting that the system sometimes failed to
initialise properly the first time it was switched on. While they were always able to
resolve this by restarting the system, School B in particular complained that this
cut significantly into the short periods they were able to find for use.
Finally, schools E, F and G all indicated that exam preparation cut into the
amount of time available for therapy, particularly in the second phase of their
deployment (the multiplayer phase for Schools E and F, single player for School
G), and that this made it particularly difficult to find time to use the system.
These comments all confirmed our initial finding from working with teachers
when developing the system in the first place, that manoeuvrability, ease of setup
and a small footprint are all important for a school environment. This demonstrates
not only the merit of the revised concept of a self-contained unit over the initial
notion of four to six independent joysticks, but also the importance of engaging
with users to properly understand their needs.
4.5 Conclusions and Further Work

This paper has reported the deployment of a novel multiplayer gaming system for
delivering arm rehabilitation in a school environment. The system has been
deployed in a real environment, and this has demonstrated that the system can be
used in a variety of schools. The games presented have been well-received despite
their simplicity although there were significant variations in individual preference
and a wider range of games would be desirable if the system were to be used in the
longer term.
There are a further three schools still to test the system before final conclusions
are drawn. In addition, this project has only assessed the feasibility of deploying
and using the system in a school environment. Clinical trials with a larger number
of children will be required to demonstrate the therapeutic efficacy of this
approach.
4.6 References
Chappell F, Williams B (2002) Rates and reasons for non-adherence to home physiotherapy
in paediatrics. Physiotherapy, 88(3): 2-11
Deutsch JE, Borbely M, Filler J, Huhn K, Guarrera-Bowlby P (2008) Use of a low-cost,
commercially available gaming console (Wii) for rehabilitation of an adolescent with
cerebral palsy. Physical Therapy, 88(10): 1196-1207
Imms C (2008) Children with cerebral palsy participate: A review of the literature.
Disability and Rehabilitation, 30(24): 1867-1884
Jackson AE, Holt RJ, Culmer PR, Makower SG, Levesley MC, Richardson RC et al. (2007)
Dual robot system for upper limb rehabilitation after stroke: the design process.
Proceedings of the Institution of Mechanical Engineers Part C: Journal of Mechanical
Engineering Science, 221(7): 845-857
Johnson A (2002) Prevalence and characteristics of children with cerebral palsy in Europe.
Developmental Medicine and Child Neurology, 44(9): 633-640
Lanningham-Foster L, Foster RC, McCrady SK, Jensen TB, Mitre N, Levine JA (2009)
Activity-promoting video games and increased energy expenditure. Journal of Pediatrics,
154(6): 819-823
Malone TW, Lepper MR (1987) making learning fun: A taxonomy of intrinsic motivations
for learning In: Snow RE, Farr MJ (eds.) Aptitude, learning and instructions, Vol. 3:
Cognitive and affective process analyses. Laurence Erlbaum Associates, Hillsdale, NJ,
US, pp 223-253
Read JC (2008) Validating the fun toolkit: An instrument for measuring children’s opinions
of technology. Cognition, Technology and Work, 10(2): 119-128
Sandlund M, Mcdonough S, Hager-Ross C (2009) Interactive computer play in
rehabilitation of children with sensorimotor disorders: A systematic review.
Developmental Medicine and Child Neurology, 51(3): 173-179
Sweetser S, Wyeth P (2005) Gameflow: A model for evaluating player enjoyment in games.
ACM Computers in Entertainment, 3(3): 3A
Weightman APH, Preston N, Levesley MC, Holt RJ, Mon-Williams M, Clarke M et al.
(2011) Home based computer-assisted upper limb exercise for young children with
cerebral palsy: A feasibility study investigating impact on motor control and functional
outcome. Journal of Rehabilitation Medicine, 43(4): 359-363
Chapter 5
Evaluating the Accessibility and Usability

of Blogging Platforms for Blind Users
B. Wentz, M. Cirba, N. Kharal, J. Moran and M. Slate
5.1 Introduction
Web-based social media have become a dynamic way of allowing the inclusive
communication of many perspectives from diverse backgrounds. In a recent survey
from WebAIM, blogging was noted to be the dominant form of social media
according to screen reader users (WebAIM, 2010). Social media and other forms of
technology can provide a platform for users with disabilities to interact and
communicate on a level playing field with anyone else throughout society. It is the
responsibility of application designers to follow established standards of
accessibility and usability in order to provide equal access to this dynamic
technology.
5.2 Related Literature

There are approximately 284 million individuals worldwide who are visually
impaired (low vision), and this figure includes 39 million individuals who are
completely blind with no residual vision (World Health Organization, 2011). A
screen reader (such as JAWS, System Access, VoiceOver, or Window-Eyes) is a
software application that reads the content of a computer screen out loud in a linear
manner, using computer-synthesised speech. This is the primary way that blind
users access computers and web sites. Braille devices are often too expensive, and
the rate of Braille literacy among blind users very low. For example, in the United
States, fewer than 10% of children who are blind are learning to read Braille
(National Federation of the Blind, 2011).
In 2005 a small study was conducted to evaluate the accessibility of several
blogging platforms (American Foundation for the Blind, 2010). It was concluded
that problems such as CAPTCHAs, which are required to register and create most
blogs, create accessibility challenges. Also highlighted were accessibility

44 Wentz et al.
challenges related to navigating many of the blog platforms. Common problems

discovered were related to basic principles of accessible design, such as properly
labeling the fields on a web-based form. Recently, WordPress has made an attempt
to provide some general accessibility guidelines for bloggers who use that platform
(WordPress.org, 2011). Our project was intended to provide a more recent
examination of a more narrow focus, namely to evaluate accessibility and usability
by enabling users who rely on screen readers to access the platforms. Our usability
testing examined the ability to read and post comments, and our accessibility
evaluation evaluated the technical ability required to create a blog and access the
main administrative console of a blog platform.
5.3 Research Methodology and Results

5.3.1 Usability Testing
5.3.1.1 Blog Platform Selection
Blogger (a Google product) and WordPress were selected for usability testing due
to suggestions that together they are the largest platforms, providing the
background structure for approximately 85% of all blogs (Peltier, 2009).
WordPress offers both WordPress.com (which provides free blog hosting and/or
paid, ad-free blogging) as well as WordPress.org (which provides blogging
software available as a download to be installed on a server for localised hosting).
Since the default templates are the same for WordPress.org and WordPress.com,
WordPress.com was selected since it is a hosted system through which any
individual is permitted to establish a free blog. The focus of the following usability
testing was on the default templates within both Blogger and WordPress, since
both platforms offer numerous templates.
5.3.1.2 Participant Selection for Usability Testing

The participants recruited for this study were self-labeled as blind, JAWS screen
reader users, and at least 18 years of age. We sent recruitment emails to the
Pennsylvania chapter of the National Federation of the Blind as well as the
Pennsylvania Council of the Blind. Additionally, the recruitment email was posted
on listservs that are commonly used by blind individuals within the state. Since
there is no central directory of all blind individuals worldwide (or in the United
States), a true random sampling would be technically impossible. A total of 15
participants were involved in the usability testing. Our goal with this evaluation
was to conduct exploratory research with a group of blind users, in order to
discover usability challenges that could be corrected to provide improved
accessibility and usability of web-based blog platforms.
5.3.1.3 Data Collection for Usability Testing

Usability testing of Blogger and WordPress was conducted in March and April
2011. The testing was conducted using an Acer Aspire One netbook with the
Windows XP operating system and Internet Explorer 8. An external keyboard,
Evaluating the Accessibility and Usability of Dominant Blogging Platforms 45
external speakers, a wireless broadband card (for testing the web-based

applications) and JAWS 12 (screen reader software) were used for the usability
testing. A stopwatch was used to record the time spent on each task. We spent time
creating test blogs and accounts that would make sure that no personal data of the
participants was used.
Each session was started by asking each participant questions about their
background and blog experience, and then participants were asked to rank the
importance of reading blog comments, posting blog comments and creating a blog.
After the initial questionnaire, we read each task to the participants, and then they
tried to complete each task without our assistance. Each task was timed and
recorded for completion, and participants could choose to stop the task at any time.
We noted problems and comments when they were mentioned by participants or
observed by us.
5.3.1.4 Demographics of Usability Testing

The participants in our study ranged in age from 34 to 63, with the mean age being
54. There was a close ratio of male (seven) to female participants (eight). Most of
the participants were college graduates or had completed some college (12 out of
15). The most common operating system was Windows XP, and the most common
Web browser was Internet Explorer 8. The versions of JAWS used ranged from
version 8 to 12, with the most common version being version 12. The mean
number of years of experience using JAWS was 10.8 years. Since the average
blind individual might have less education and screen reader experience than the
participants in this study, it is likely that they would have more difficulty with
these blog interfaces than the study participants did.
5.3.1.5 Previous Experience and Value of Blogging

When asked about their previous experience with blogs, 80% (12 out of 15) of the
participants reported having previously read comments that were posted on a blog,
and out of those participants, 47% (7 out of 15) recalled having no difficulty when
reading blog comments. Participants were then asked about posting comments on a
blog, and only 47% (7 out of 15) reported previously posting comments on blogs.
Of those participants, 57% (8 out of 15) recalled having difficulty when posting
blog comments. Only 27% (4 out of 15) of the participants reported previously
creating or attempting to create a blog, and out of those, two of the four reported
having difficulty when trying to create a blog.
Participants were then asked three questions related to the value that they place
on being able to read comments, post comments, and create blogs. On a scale of 1-
5 (with 5 being the most important), the mean responses were very close for all
three. Reading comments had a mean of 3.5, posting comments had a mean of 3.7,
and creating blogs had a mean of 3.5. This indicates that for this sample, a
moderate value is placed on all aspects of interacting with blogs.
5.3.1.6 Testing Results and Impact of Previous Experience

After the preliminary questions about demographics and prior experience with
blogs, users were then asked to complete two tasks in both Blogger and
WordPress. For Task 1, participants were asked to read any comment posted on the
46 Wentz et al.
blog web page. For Task 2, participants were asked to post a comment about the
article on the blog web page. The tasks with the results are illustrated in Table 5.1.
Blogger had a higher rate of completion for Task 1, but the task took longer to
complete than it did in WordPress. No participants were able to complete Task 2 in
Blogger due to a usability problem which is described later.
Table 5.1. Tasks and the completion rate, mean time, and SD for successful tasks
Task Blog Completion Mean Standard
Platform Percentage Time Deviation
1: Read a comment Blogger 80% (12 out of 15) 144.3 sec. 75.2 sec.
1: Read a comment WordPress 67% (10 out of 15) 126.8 sec. 36.1 sec.
2: Post a comment Blogger 0% - -
2: Post a comment WordPress 53% (8 out of 15) 129 sec. 54.3 sec.
The impact that previous experience with reading and posting comments had
on the results of this usability testing is illustrated in Table 5.2. It is important to
note that this does not reflect previous experience with a particular interface but
rather previous experience with either reading a blog comment or posting a blog
comment on any blog interface. For task 1 on the Blogger interface, it appears that
there is an improvement in usability based on prior familiarity, but it is important
to note that there were only a small number of users (three) who were
inexperienced with task 1. Task 2 in WordPress does seem to indicate a usability
curve based on experience, since the number of experienced to inexperienced users
was much closer, yet the task completion rate was heavily weighted towards
experienced users. In addition, the mean time to complete successful tasks was
higher for inexperienced users for Blogger task 1, and for WordPress tasks 1 and 2.
Table 5.2. Previous experience and mean completion rate/time for successful tasks
Task/Interface Users with Experienced Inexperienced
Experience Users Users
Task 1, Blogger 12 out of 15 92% (11/12) 33%. (1/3)
135.6 sec. 217.7 sec.
Task 1, WordPress 12 out of 15 67% (8/12) 67% (2/3)
119.6 sec. 155.5 sec.
Task 2, Blogger 7 out of 15 0% 0%
- -
Task 2, WordPress 7 out of 15 71% (5/7) 38% (3/8)
113 sec. 155.7 sec.
5.3.1.7 Usability Problems Identified

While many minor usability challenges were identified during the usability testing
conducted for this study, there were several major usability challenges which, if
corrected, could dramatically improve the usability of these blog interfaces.
Usability problems with Blogger included a “Post Comment” button that did not
work as expected and an audio CAPTCHA that is not embedded and requires an
external plug-in to work. These problems effectively prevent users from
successfully posting comments.
Figure 5.1. Post comment problem in blogger
In Figure 5.1, there is a screen shot of the “Post a Comment” form for the
Blogger interface. While this should be a simple, straightforward process, when a
user (blind or sighted) selects the “Post Comment” button, the action is not
performed, and instead the user is redirected back to the same page, and the
comment is not posted. It takes several attempts before the user is directed to the
login prompt for account credentials. This does not occur if a user is already
logged into a Google Account. While this is frustrating for any user, the problem is
not immediately evident to a blind user. The only indication that this is occurring is
the small red image that is shown in the screenshot above. One user, who tried this
process repeatedly and did reach the next step, was taken to a prompt for a
CAPTCHA. Figure 5.2 illustrates the traditional CAPTCHA screenshot for the
Blogger interface with a symbol indicating an audio option for accessibility. The
alternate text for the audio option says “Listen and type the words you hear.” The
problem is that the user was unable to proceed because the CAPTCHA did not rely
on embedded audio but on a particular format of audio which required a plug-in for
use.
Figure 5.2. Audio CAPTCHA plug-in problem in blogger

48 Wentz et al.
Usability problems with WordPress included a link that directs users to the
wrong field in a form, and an error message that could be improved for usability.
When a user selects the reply link to add a comment, the focus of the cursor goes to
the “Comment” field on the form. For a sighted user, it would be evident that
“Name” and “Email” are required fields, but for a user relying on a screen reader
and keyboard navigation, the usability problem with the “Reply” link is confusing
and could prevent a successful post unless the user realises what the problem is.
5.3.2 Accessibility Evaluations

Accessibility evaluations often rely on automated software (such as A-Prompt,
Deque WorldSpace, and WAVE). Automated evaluations can determine that a web
page generally fails compliance with standards such as Section 508, but automated
evaluations are often not as accurate as manual evaluations because accessibility
problems such as alternate content descriptions that do not fit the context cannot be
currently determined by automated means. Careful inspections using a screen
reader (such as JAWS) are considered to be the most accurate form of accessibility
evaluation (Mankoff et al., 2005), and that accuracy increases when multiple
individuals evaluate the same interfaces (Lazar et al., 2010a). This process can be
used to determine compliance with standards, such as the Section 508 standards for
web sites. Accessibility evaluations complement usability evaluations in that
usability evaluations can assess ease of use, whereas usability testing can often not
completely evaluate compliance with standards since users may not be able to
completely use all aspects of a system due to accessibility barriers.
5.3.2.1 Procedures for the Accessibility Evaluations

The evaluations that we conducted utilised the web site accessibility guidelines of
Section 508 (1194.22) of the US Rehabilitation Act, identified as paragraphs “a”
through paragraph “p” (US Government, 2010). This inspection process involves
an individual (previously trained on accessibility inspections) with vision
inspecting a web site using a screen reader and a checking for compliance with
each guideline from Section 508. Table 5.3 presents a list of the guidelines, along
with a short description of each of the guidelines (note that the descriptions are
summarised from Lazar et al., 2010a, and not a part of the Section 508 guidelines).
Table 5.3. Description of each of the 16 paragraphs of the Section 508 web guidelines
(a) Text Equivalent (have a text equivalent for any graphical elements)
(b) Synchronised Equivalent Alternatives (have captioned video, transcripts of any audio, or
other alternatives for multimedia)
(c) Use of Color (color should not be used as the only method for identifying elements of the
web page or any data)
(d) Organisation (style sheets are encouraged, but users should still be able to utilise a web
page when style sheets are turned off)
(e) Redundant Text Links on Server-Side Image Map and (f) Client-Side Image Maps
(redundant clickable links for server-side image maps, and accessible client-side image
maps are preferred)
(g) and (h) Row and Column Headers (use appropriate headers and markup to allow easy
navigation of a table)
(i) Frames (title all frames and label all frames for easy identification and navigation, e.g.,
use “navigation” “main content” and “search” rather than “top” or “bottom”)
(j) Screen Flicker Frequency (limit or eliminate the use of flickering, which can provoke
seizures)
(k) Text-Only Page Default (if a web page cannot be made accessible, provide an equivalent
text-only page, and make sure it is kept up to date)
(l) Scripting Languages (make sure that equivalents for any non-accessible scripting are
included, e.g., for those who are not using pointing devices)
(m) Linked Plug-In or Applet (if any plug-ins are required, make sure to provide a link to an
accessible version of the plug-in)
(n) Online Electronic Forms (all forms must be properly labeled and accessible)
(o) Method to Skip Repetitive Navigation Links (all web pages should have a link which
allows a user to skip directly to the main content, bypassing any site navigation information)
(p) Alerts on Timed Responses (if any page responses are timed, the user should be given
the opportunity to indicate that more time is needed)
For this evaluation, we selected the pages used to create a blog as well as the
main interface for managing the blog in both Blogger and WordPress (refer to
Table 5.4).
Table 5.4. Web pages on blogger and WordPress selected for evaluation
Blogger.com Web Pages: WordPress.com Web Pages:
Home page Home page
“Get started” page to create an account Page to create an account
Page to name the blog Blog management page
Page to choose the template
Blog management page
For each web page, four individuals did a separate inspection using a screen
reader, followed-up by a code inspection. After doing an individual inspection, the
four individuals met, discussed the differences between their individual
evaluations, interpreted, re-inspected the pages, and then agreed upon one common
50 Wentz et al.
evaluation for each web page. This is a common approach to provide a higher level
of validity than a single individual review (Nielson and Mack, 1994).
5.3.2.2 Results of the Accessibility Evaluations

The accessibility evaluation revealed some major accessibility barriers with the
registration process and the management interface for both Blogger and
WordPress. Each page that was evaluated had at least some minor accessibility
violations. Some of the violations for Blogger included images with no alternate
text (paragraph “a”), tables with headers that were missing mark-up (paragraph
“h”), labels that were not properly associated with their controls, and poor error
labeling (paragraph “n”). Blogger also violated paragraph “c” by using color alone
to identify information on the “Get Started” page. During the blog template
selection process, paragraph “l” was violated by the template selection feature
being inaccessible with the keyboard alone. On the main page to manage the blog,
Blogger violated paragraph “b” (with no caption for the video on how to use
Blogger), paragraph “g” (with no headers in a table), and paragraph “l” (drop-down
selections for text formatting were created using inaccessibly implemented
JavaScript).
On WordPress, the home page was missing alternate text for images (paragraph
“a”), and a search form field was not properly labeled (paragraph “n”). On the
sign-up page, color alone was used to convey meaning (paragraph “c”), there was a
table without header information (paragraph “h”), there was a scripted drop-down
list on a form that was not accessible (paragraph “l”), and there was another search
form field that was not labeled (paragraph “n”). There were many violations on the
main blog management page (7 out of 16 paragraphs), including missing alternate
text (paragraph “a”), no captions for the tutorial video (paragraph “b”), missing
headers for tables (paragraph “h”), untitled frames (paragraph “i”), a JavaScript
“admin” bar that was not accessible (paragraph “l”), and no way to skip over
navigational links to reach the main content (paragraph “o”). Figure 5.3 shows a
screenshot of the JavaScript “admin” bar on the main blog management page
which was only accessible by using a mouse.
Figure 5.3. Inaccessible “admin” bar in WordPress

5.4 Discussion
The major usability problems discovered during the usability testing process of the
Blogger and WordPress interfaces are primarily simple issues that could easily be
corrected by developers and could have been avoided through some basic usability
evaluations of these interfaces. As far as the audio CAPTCHA plug-in problem
with Blogger, it is ironic that Google owns both Blogger and reCAPTCHA, which
is a company that develops CAPTCHAs for commercial use. The audio
CAPTCHAs produced by reCAPTCHA use embedded audio, and while they may
have usability problems related to audio CAPTCHAs in general (Lazar et al.,
2010b), at least the external plug-in to play the CAPTCHA could be avoided.
The accessibility problems identified during the accessibility evaluations of the
Blogger and WordPress registration and management processes were also
problems that could easily be corrected. Adding alternate text, adding labels to
form fields, adding skip navigation links, and adding headers to tables are all very
simple tasks for web designers. A basic accessibility evaluation would reveal these
problems, and evaluations at regular intervals could prevent problems like these
from occurring in the future.
Companies who provide products such as Blogger and WordPress, and anyone
who uses these products should be aware of these accessibility and usability
problems. The impact of web-based blogs is far-reaching, and simply paying
attention to the design of these common templates could significantly improve the
online experience of millions of users.
5.5 Conclusions
A few of the problems uncovered in this study could affect both blind and sighted
users to some extent (such as the “Post Comment” problem in WordPress):
however, navigating with only a keyboard and screen reader causes many of these
problems to become significant. Designers should carefully follow national
guidelines such as those set forth in US Section 508 and international guidelines
such as the Web Content Accessibility Guidelines. Regular accessibility
evaluations and usability testing would ensure that these popular blogging
platforms are equally accessible to all users.
5.6 References
American Foundation for the Blind (2010) Is blogging accessible to people with vision loss?
Available at: http://www.afb.org/Section.asp?SectionID=57&DocumentID=2753 (Accessed 13
December 2010)
Lazar J, Beavan P, Brown J, Coffey D, Nolf B, Poole R et al. (2010a). Investigating the
accessibility of state government web sites in Maryland. In: Langdon PM, Clarkson PJ,
Robinson P (eds.) Designing inclusive interactions. Springer, London, UK, pp 69-78
52 Wentz et al.
Lazar J, Feng J, Adelegan O, Giller A, Hardsock A, Horney R et al. (2010b) Assessing the
usability of the new radio clip-based human interaction proofs. In: Poster Presentation at
the 6th Symposium on Usable Privacy and Security (SOUPS 2010), Redmond, WA, US
Mankoff J, Fait H, Tran T (2005) Is your web page accessible? A comparative study of
methods for assessing web page accessibility for the blind. In: Proceedings of the 23rd
ACM Conference on Human Factors in Computing Systems, Portland, OR, US, pp 41-50
National Federation of the Blind (2011) How many children in America are not taught to
read? Available at: http://www.nfb.org/nfb/braille_initiative.asp (Accessed 5 May 2011)
Nielson J, Mack R (1994) Usability inspection methods. John Wiley and Sons, NY, US
Peltier J (2009) Which blogging platform do you use? Available at:
http://peltiertech.com/WordPress/which-blogging-platform-do-you-use/ (Accessed 5
December 2010)
US Government (2010) Section 508 standards guide. Available at: http://www.section508.gov/
index.cfm?fuseAction=stdsdoc#Web (Accessed 5 May 2011)
WebAIM (2010) Screen reader user survey #3 results. Available at:
http://webaim.org/projects/screenreadersurvey3/ (Accessed 5 January 2011)
Accessibility (2011) WordPress.org. Available at: http://codex.WordPress.org/Accessibility
(Accessed 15 January 2011)
World Health Organization (2011) Visual impairment and blindness. Available at:
http://www.who.int/mediacentre/factsheets/fs282/en/ (Accessed 5 May 2011)
Part II
Measuring Demand
and Capabilities
Chapter 6
A Population Perspective on Mobile Phone

Related Tasks
M. Bradley, S. Waller, J. Goodman-Deane,
I. Hosking, R. Tenneti, P.M. Langdon and
P.J. Clarkson
6.1 Introduction
For design to be truly inclusive, it needs to take into account the range of users’
capabilities. To do this appropriately, good data on those capabilities is needed.
This paper reports on results from a postcode sampled survey of 362 people. The
survey examined a wide range of user capabilities and characteristics, but the paper
focuses on just a few of the survey measures. These measures examine some of the
component activities involved in using mobile telephones: selection of a menu item
via two different interaction patterns, use of differing sized pushbutton controls and
insertion of two different types of electrical connector. These results can help to
inform more inclusive design of mobile phones by examining how people’s
capability to perform these activities varies across different activities and by age
and gender.
The survey from which this paper draws its data, aimed to test methods and
materials in preparation for a follow-up survey with a bigger sample. Although the
data in this paper was taken from a preparatory survey, there were 362 participants,
and this sample size is big enough to achieve statistical significance for the
conclusions drawn.
Further details of how the survey was designed and conducted are now
presented, followed by details of the specific tests that are reported here. Related
research is also presented for each specific test.
6.2 Method
The survey examined a wide range of human capabilities and characteristics related to
product use. It was part of the i~design research programme (i~design, 2011). It was
designed by the i~design research team, and conducted by the National Centre for

56 Bradley et al.
Social Research (NatCen). NatCen is a non-commercial professional survey

organisation. 30 professional interviewers from NatCen conducted the actual surveys,
and each interviewer received one day’s training supervised by the i~design team.
6.2.1 Sample
The sample was recruited through random selection of private-household postcodes in
England and Wales. Invitation letters were sent out to 990 postcode addresses, and each
property was subsequently visited by one of the interviewers. At households that
consented to take part, the interviewer selected a single occupant at random from those
aged 16 and over, and the participants were not paid. 362 responses were obtained, and
53.6% of these were female. 23% of the sample were aged 16-34, 29% were aged 35-
49, 24% were aged 50-64 and 23% were aged over 65.
6.2.2 Survey Procedure

The survey was conducted face-to-face using a computer assisted personal interviewing
(CAPI) programme on a laptop.
The survey examined a wide range of areas of human capability and product use:
vision, hearing, dexterity, mobility, reach, cognitive function, technology/product use,
psychological resources, and anthropometrics. The survey aimed to gather generic
information that could be used to predict participants' ability to interact with products.
Some of the tests were therefore designed to elicit information on participants’ basic
capabilities, which are relevant to many aspects of product interaction.
For example, vision tests measured participants’ visual acuity, which affects
performance at a range of product tasks, such as reading text on a mobile phone screen
and reading a manual. However, the survey also included some specific product tasks
that require a combination of basic capabilities. For example, the ability to press
particular buttons on a mobile phone was measured directly, because this typically
requires hand-eye coordination and fingertip sensitivity.
This paper focuses on the product tasks in the survey that relate specifically to
mobile telephone use: selection of a menu item via two different interaction patterns,
use of differing sized pushbutton controls and insertion of two different types of
electrical connector. These measures are described in more detail below. Many of the
other measures in the survey are also relevant to mobile phone use and design. The full
results from the survey and its documentation are publicly available: please contact
edc-enquiries@eng.cam.ac.uk for details.
6.2.3 Menu Selection

As part of the survey, participants were presented with two interface styles for selecting
an item from a list. We use the term ‘select and confirm’ to refer to an interface in
which users navigate up and down the list using arrow keys and press a ‘select’ key
when they reach their choice (Figure 6.1a). Similarly, ‘number navigation’ refers to an
interface in which users choose an item by entering the number next to it (Figure 6.1b).
A Population Perspective on Mobile Phone Related Tasks 57
(a) (b)
Figure 6.1. Mock-up paper prototype mobile phone interfaces used to assess (a) Select and
confirm (b) Number navigation interaction styles
The ‘select and confirm’ style is more common in mobile telephones, but previous
studies have indicated that some users experience problems with it. Lindholm and
Keinonen (2003) describe how novice users can initially struggle with the concept of
soft keys in interfaces of this type but this can be overcome through practice. The
widespread adoption of this interface style for technology products means that many
younger people are familiar with it, but many older users may still have limited or no
experience with it. These older users are therefore more likely to struggle with this
menu style. This difficulty was specifically observed for older users in digital set-boxes
for television (Clarkson and Keates, 2003), which often have ‘select and confirm’
menus. Clarkson and Keates recommended the use of number lists as an alternative
approach.
Our study thus investigated whether the ‘number navigation’ approach could
overcome the problems that some users experience with ‘select and confirm’. We
expected that younger users would be experienced and proficient at using ‘select and
confirm’ menus and that many older users (aged 65+) would not. The more direct
interaction of ‘number navigation’, which requires only one key press to select a menu
item, could be advantageous for such novices.
These interaction styles were presented to users using a simplified paper
prototyping method (Snyder, 2003). Respondents were shown images of mobile
telephones with these interfaces (Figure 6.1), and were asked to indicate which buttons
they would press to see what information is in the calendar. The order in which the
interfaces were presented was counter-balanced. The interviewer scored whether the
correct buttons were pressed in the correct order.
The ‘number navigation’ interface requires fewer physical buttons to implement
than the ‘select and confirm’ interface, and this reduction of complexity represents a
confounding variable. However, the decision was made to present both types of user
interface with the minimum number of buttons required to implement basic phone
functionality. If this test revealed a useful performance difference, this could offer real-
world benefit for the user.
The test was limited by using a simplified paper prototyping method, where the
participant was not given any feedback on button presses. This was necessitated by
practical constraints; the survey was conducted by 30 different interviewers and with
only one day of training to cover the whole survey, it was not feasible to train them to
58 Bradley et al.
reliably run any type of interactive prototyping. This limitation should be taken into
account in interpreting the results, as failure rates on the tasks are likely to be higher
without feedback to the user. However, the method was intended to measure whether,
on first sight of an interface, the user’s first attempt to plan and execute the task would
involve any incorrect or unnecessary actions. Even if a user could eventually succeed in
a goal through trial, error and recovery, an interface that allows users to get it right first
time is preferable.
Another factor that should be taken into consideration is that ‘number navigation’
requires fewer key presses, and thus presents fewer opportunities for the user to make a
mistake. This is a potential confounding variable which is hard to eliminate in a simple
experiment of this type. The decision was made to present the user interfaces in a form
that is likely to be used in practice. This has the advantage that, if there is a useful
performance difference, then the improved interface can be incorporated in an actual
mobile phone, even if there is potential uncertainty about the underlying reason for the
benefit.
6.2.4 Pressing Buttons

The survey also assessed participants’ ability to use differing sized pushbutton controls
on a mobile telephone. The participants were presented with an LG KP170 mobile
phone with a small five way keypad and four larger surrounding pushbuttons (Figure
6.2). They were shown the five way keypad and asked to press four of the keypad
arrows, in the order of up, right, down and centre (‘OK’ button). They were then asked
to press the four larger outer buttons in the order of bottom left, top left, top right and
bottom right.
After each task, the interviewer recorded whether the correct screen was shown. For
the first task, this indicated whether the respondent had pressed the buttons correctly in
the specified order. However, an experimental weakness meant that, for the second
task, this only indicated whether the last of the buttons had been pressed correctly. Any
participant who tried the task was also asked to rate how easy they found the task on a
six point scale from very easy to impossible. They were not given any feedback on
whether they had successfully completed the task before making this assessment.
We expected participants (particularly older participants) to perform better with
the large buttons, in accordance with Fitts’ Law (Fitts, 1954), because they require
less accurate finger positioning and dexterity.
Large
Five way outer
keypad buttons
Figure 6.2. LG KP170 mobile phone showing the buttons used in the study
6.2.5 Inserting Cables

Participants were also tested on their ability to insert two different types of electrical
connector. Participants inserted two connectors into an MP3 player (a Sansa Clip+): a
3.5mm headphone jack plug and a mini-USB plug (Figure 6.3). An MP3 player was
used for purposes of experimental convenience, but it was felt that the task of insertion
would not differ from that experienced on a mobile phone.
Firstly, the MP3 player and headphone jack socket location were shown to the
participants. They were then given the MP3 player and the headphone jack plug and
were asked to insert the jack into the socket. The test was aborted if the participant took
longer than 15 seconds. The participants were then shown the mini-USB socket
location on the MP3 player, and the mini-USB jack (at a random but incorrect
rotational orientation). They were handed the MP3 player the right way up with the
mini-USB jack on top of it with the mini-USB symbol visible, and were asked to insert
the jack in the indicated socket.
After each task, the interviewer recorded whether the jack was inserted correctly.
Participants who did the task successfully also rated how easy they found the task on a
six point scale from very easy to very difficult.
Figure 6.3. MP3 player showing headphone jack and mini-USB sockets and plugs
We expected participants to perform better with the headphone jack than with the
mini-USB cable because the jack on the mini-USB needs to be rotated to the correct
orientation before insertion, but the headphone jack can be inserted in any orientation.
6.3 Results and Discussion

6.3.1 Menu Selection Results
Overall, more people completed the ‘select and confirm’ task successfully than the
‘number navigation task’ (p<0.001, McNemar’s). For each of the tasks, successful task
completion decreased with age (p<0.001, Pearson’s chi-squared) as shown in Figure
6.4. This was most evident in the two oldest groups, particularly for women in the 75+
group. For both tasks, there was no significant difference between men and women
(p>0.05, Fisher’s exact).
60 Bradley et al.
As explained in Section 6.2.3, we expected that younger users would perform better
with the ‘select and confirm’ interface than older users. This was supported by the
results, with 72% of people under 65 using ‘select and confirm’ successfully, and 61%
using ‘number navigation’ successfully (p<0.001, McNemar’s). We also expected that
older users (65+) would perform better with the ‘number navigation’ style than with
‘select and confirm’. The results indicate that 28% of over 65s used ‘select and
confirm’ successfully, and 33% used ‘number navigation’ successfully, but this was not
statistically significant (p>0.05, McNemar’s), possibly due to only having 38 people in
this age group. Further analysis and research is needed to better understand how
different kinds of participants use different interface styles.
Select and confirm' menu task 'Number navigation' menu task

100
% of age band correct
80
60
40
20
0
18-24
25-34
35-44
45-54
55-64
65-74
18-24
25-34
35-44
45-54
55-64
65-74
75+
75+
age band age band
Male Female Male Female
Figure 6.4. Percentage of male and female participants succeeding in the menu selection
tasks, by age category
As explained in Section 6.2.3, we expected that younger users would perform better
with the ‘select and confirm’ interface than older users. This was supported by the
results, with 72% of people under 65 using ‘select and confirm’ successfully, and 61%
using ‘number navigation’ successfully (p<0.001, McNemar’s). We also expected that
older users (65+) would perform better with the ‘number navigation’ style than with
‘select and confirm’. The results indicate that 28% of over 65s used ‘select and
confirm’ successfully, and 33% used ‘number navigation’ successfully, but this was not
statistically significant (p>0.05, McNemar’s), possibly due to only having 38 people in
this age group. Further analysis and research is needed to better understand how
different kinds of participants use different interface styles.
In fact, the success rates for both menu selection styles were low overall, with only
61.8% of the sample using the ‘number navigation’ style successfully, and 73.4%
succeeding with ‘select and confirm’. These rates are even lower for older age groups,
with over 50% of participants aged over 75 failing on both tasks. It should be
remembered that the partial paper prototyping method used is likely to give lower
success rates than use in practice. Nevertheless, the success rates are sufficiently low to
raise serious concerns about both menu selection styles. It seems that, although number
navigation may perform slightly better with older people, it is not effective enough to
work as an alternative. Other interface styles for selecting menu options are needed,
and more work is required. Some possibilities are provided by touchscreen interfaces,
which allow menu options to be selected more directly. However, this technology may
have other usability issues including those arising from the lack of tactile feedback.
6.3.2. Button Dexterity Results

Overall, more people were successful with the larger buttons than with the fiveway
keypad (p<0.001, McNemar’s), in agreement with our hypothesis. However, an
experimental weakness emerged for this condition alone: participants only needed to
press the last button correctly to be recorded as having completed the task correctly,
thus potentially improving the success rate. For the fiveway keypad, participants
needed to press all the buttons correctly to be recorded as completing the task
successfully.
Thus, it may be more useful to examine the results for each task separately. For the
fiveway keypad (Figure 6.5a), there was no significant difference between men and
women, but success on this task did decrease with age (p<0.001, Pearson’s chi-
squared), with 47% of those aged 75+ failing on this task. For the larger outer buttons
(Figure 6.5b), there was no significant difference between men and women or between
different age groups, with high levels of success across the sample.
Participants also rated how easy they found each task (see Figure 6.6). Considering
all ages and genders together, participants rated the larger buttons as easier to use than
the small ones (p<0.001, Marginal Homogeneity). For the fiveway keypad, it seems
that the ease of use ratings may decrease with age but increase again for the oldest age
group, but further analysis and investigation is needed.
5 way keypad (success) Bigger buttons (success)
100
80
% of age band
60
40
20
0
18-24
25-34
35-44
45-54
55-64
65-74
18-24
25-34
35-44
45-54
55-64
65-74
75+
75+
age band age band
(a) (b)
Figure 6.5. Percentage of male and female participants succeeding in the button pressing
6.3.3 Cable Insertion Results

Overall, people were more successful at inserting the mini USB cable than the
headphone jack (p<0.001, McNemar’s). This contradicted our expectation that the
headphone jack would perform better. Further examination of the data indicated that
many people failed because they inserted the jack partially but not fully. It is likely that
many of these partial insertions were because the required pushing force peaked at a
62 Bradley et al.
partial insertion point. Some participants may have believed that they had completed
the task at that point and not thought to push the jack harder to insert it fully. This
appeared to be more problematic for the female participants (p<0.001, Fisher’s exact),
but it is unclear why. This may be due to muscle strength differences or a difference in
prior experience such that participants did not know that the plug needed to be inserted
further.
Other age and gender differences are shown in Figure 6.7. Success on the
headphone jack task decreased with age (p<0.01, Pearson’s chi-squared), as did success
on the mini-USB task (p<0.05, Pearson’s chi-squared), although the latter may be
affected by ceiling effects. There was no significant difference between genders for the
mini-USB task, with almost all participants successfully completing this task.
5 way keypad (easy or v. easy) Bigger buttons (easy or v. easy)

100
80
% of age band
60
40
20
0
18-24
25-34
35-44
45-54
55-64
65-74
18-24
25-34
35-44
45-54
55-64
65-74
75+
75+
age band age band
Figure 6.6. Percentage of male and female participants rating the task as ‘very easy’ or
‘easy’ for the five way keypad (small) and outer large button tasks, by age category
Headphone jack (success) Mini USB (success)

100
80
% of age band
60
40
20
0
18-24
25-34
35-44
45-54
55-64
65-74
18-24
25-34
35-44
45-54
55-64
65-74
75+
75+
age band age band
Figure 6.7. Percentage of male and female participants succeeding in the cable insertion
Participants also rated how easy they found each task (see Figure 6.8). Although
almost all (over 99% of) participants completed the mini-USB task successfully, they
considered it to be more difficult than the headphone jack task (p<0.001, Marginal
Homogeneity). This agrees with our expectation that this task would be more difficult.
It may be because participants needed to orient the plug with the socket in this task.
Headphone jack (easy or v. easy) Mini USB (easy or v. easy)

100
80
% of age band
60
40
20
18-24
25-34
35-44
45-54
55-64
65-74
18-24
25-34
35-44
45-54
55-64
65-74
75+
75+ age band
age band
Figure 6.8. Percentage of male and female participants rating the task as ‘very easy’ or
‘easy’ for the cable insertion tasks, by age category
6.4. Conclusions
This study examined 362 participants’ performance in and ratings of ease of use for
some component activities of mobile phone use. Across the activities, older users had
markedly lower success rates and generally lower ratings for ease of use.
In particular, the menu selection methods tested in the study had low success rates
overall and particularly for older users. This raises concerns about the suitability of these
common interface styles, and highlights the need for further research to develop new
methods of selecting menu items on mobile devices. In addition, the button pressing
tasks indicated that small buttons in the fiveway keypad were difficult for older users to
press accurately. Yet buttons of this size and type are commonly used in mobile
telephones. Somewhat, surprisingly, many participants also struggled with correctly
inserting the headphone jack. Although most people did successfully insert the mini-
USB cable, many people did not consider this task to be easy.
Overall, these results indicate that current mobile phone designs are not meeting the
needs and capabilities of older users. More detailed analysis and study are needed to
illuminate some of the fundamental issues, and further work is needed to produce more
inclusive mobile phone designs.
The survey was conducted as part of the inclusive design research programme (i~design,
2011), funded by EPSRC. As well as the authors of this paper, several others made
substantial contributions to the design and analysis of this survey, including Felicia
Huppert, Kai Ruggeri, Eddy Elton, Jose Liht and John Ryan. Mike Bradley’s time was
partly funded via the RCUK Digital Economy Programme’s BRIDGE project.
64 Bradley et al.
6.6 References
Clarkson PJ, Keates S (2003) Digital television for all. A report on usability and accessible
design. Appendix E - investigating the inclusivity of digital television set-top receivers.
Department of Trade and Industry, UK
Fitts PM (1954) The information capacity of the human motor system in controlling the
amplitude of movement. Journal of Experimental Psychology, 47(6): 381-391
i~design (2011) Available at: www-edc.eng.cam.ac.uk/idesign3/ (Accessed 31 August 2011)
Lindholm C, Keinonen T (2003) Mobile usability: How Nokia changed the face of the mobile
phone. McGraw-Hill, New York, US
Snyder C (2003) Paper prototyping: The fast and easy way to design and refine user interfaces.
Morgan Kaufmann Publishers, San Francisco, US
Chapter 7
How to Use Virtual and Augmented Reality

Techniques to Design Highly Usable
Human-machine Interfaces
S. Ceccacci, M. Germani and M. Mengoni
7.1 Introduction
Modern technological products, working and living environments, are ever more
rich in potential functionalities for end-users (e.g. communication support,
facilitation of physical actions, etc.). The increasing functionalities make these
products complex to develop. Research efforts are focused on the development of
highly usable products and environments that should be easily and intuitively used
by a wide range of people (children, elderly, disabled, etc.). User-centred design
approaches have to be adopted. This is not easy to handle when complex
environments populated by different products have to be conceived, analysed and
evaluated. In this scenario, the aim of the present work is to study a user-centred
design method in relation to the kitchen environment. It has been developed in the
context of an Italian National Research Programme, called “e-kitchen: smart and
highly usable kitchen”, started in April 2011. The project involves 16 important
Italian and international companies (e.g. Indesit, Lube Industries, Telecom Italia,
Faber-Franke), seven SMEs and five research centres, under the coordination of
Università Politecnica delle Marche. The project aims to achieve a complete
rethinking of the kitchen environment taking into account safety, comfort, eco-
sustainability and energy efficiency. One of the main goals is the improvement of
its usability for aged and disabled people. In this context it is essential to manage
and optimise the interaction between the user and the different devices so as to let
the individual live in an environment offering well-being, safety and new
intelligent functionalities. The interaction studied involves three elements: human,
machines and environment.
Multi-sensory Virtual Reality and Augmented Reality techniques are used to
support the inclusive design of all target devices and of the kitchen environment as
a whole. These technologies give the advantage of simulating user interaction from
the first conceptual design stages and hence of evaluating different design

66 Ceccacci et al.
alternatives by minimising development cost and time. Moreover, their high

usability allows the involvement of customers in the entire design process thus
implementing a user-centred design approach. In this way it is possible to achieve
customised products by directly analysing different design solutions with end-users
and finding which one meets their needs better. AR/VR technologies allow the
assessment of both physical and cognitive ergonomics. Specific protocol analysis
methods are under development in order to examine the implicit and explicit user
behaviour.
7.2 Which Supporting Prototyping Technique Fits

the Requirements of Designing Interactive
Devices?
In order to design highly usable interactive systems, it is crucial to understand the
needs and limitations of end users. This can be achieved by adopting a user-centred
perspective at each stage of the design process. According to ISO 13407 (1999)
standard, a proper User-Centred Design (UCD) process is structured in four
iterative phases. A critical one involves the building of interactive prototypes to
support rapid product usability evaluation (Mengoni and Peruzzini, 2010).
Two main prototyping techniques, which differ in the level of fidelity and
interaction they provide, can be classified. Low-fidelity prototypes (e.g. paper
sketches, cardboard mock-up) are good to test aspects such as the layout of
controls. However, they do not allow the evaluation of the effects of tactile,
auditory and visual feedback (Hall, 2001). High-fidelity prototypes (e.g. software-
based, physical mock-up) allow users to realistically appraise product aesthetic
attributes and functionalities (Sonderegger and Sauer, 2010), but they are
expensive and cannot be realised during the first conceptual design stage when the
product design has not yet been completed.
To reduce the gap between low and high fidelity prototypes, Virtual Reality
(VR)-based technologies have been introduced to create, manipulate and explore
virtual prototypes as well as to simulate product behaviour in different working
conditions. They aim to replace physical mock-ups with virtual ones (Wilson and
D’Cruz, 2006). Some studies demonstrate how they can be used to rapidly carry
out usability testing, to reduce evaluation time and costs and to involve end-users
from the earliest stages of the design process (Kuutti et al., 2001). However, VR
environments often show multiple technological limitations, such as low sense of
immersion, poor physical interaction, high complexity, low realism, unnatural
manipulation, intrusiveness and non-intuitiveness.
Mixed Reality (MR) environments represent a compromise solution in which
real and virtual worlds are combined in various proportions and presented as a
unified whole. MR fuses the two extremes of the Virtuality-Reality continuum by
synchronising information from the digital space and the physical one in a more
natural way (Milgram et al., 1994). Within the MR framework the Augmented
Reality (AR) technique is one of the most adopted due to the low cost of the
How to Use Virtual and Augmented Reality Techniques 67
technologies and to its ability to enhance the real scene with computer graphics and
emerging tactile and auditory displays (Zhou et al., 2004).
Many different solutions have been proposed in literature with the intent of
providing devices to interact with the AR environment in a more intuitive way.
Some researchers integrate handheld haptic technologies to reproduce the real
contact with objects during the exploration of virtual space (Bordegoni et al.,
2009). Others try to eliminate the gap between the interaction with a natural
environment and the interaction with a computer system by adopting Tangible
Augmented Reality (TAR) techniques (Park et al., 2008).
A lot of studies have been carried out in order to analyse how VR interfaces are
able to support elderly people to improve rehabilitation involvement (Flores et al.,
2008; Alankus et al., 2010). These studies can be useful for the development of
products and services tailored to elderly people with cognitive issues, in order to
support them in their daily lives (Pittarello and De Faveri, 2006; Kim and Dey,
2009). However, no studies have been conducted yet to assess what the optimum
VR technology is for the involvement of elderly people in design solutions
evaluation. Most of the projects which aim to address the relationship between
older people and technology, including PERSONA (2007) and UTOPIA (2011),
have been limited in their development of a new methodological approach only to
design for elderly people, without taking into consideration design-oriented
technologies issues. However, this aspect is becoming highly important to really
implement a UCD approach whilst reducing development costs. This is
demonstrated by two integrated projects of the 7th Framework Programme, which
are VAALID (2011) and VERITAS (2011).
7.3 The Proposed UCD Approach Based on VR/AR

Techniques
User involvement is particularly important in inclusive design because designers
have very different awareness of elderly and disabled users’ needs. However, it can
be a complicated task. The general proposed design process described below
(Figure 7.1) is based on a proper user-centred methodology where user
characteristics are the main drivers of the design solution identification.
The information which is required to feed into the design process is grouped
into two areas. The first includes all the information about end-user characteristics
and abilities (characteristics of elderly people) and environmental and social
factors which can affect their activities in their daily life. The information collected
into the second area is related to the characteristics of suitable technologies (e.g.
cost, flexibility, extendibility, upgradeability, etc.) and to the respective potentially
valuable applications (e.g. safety monitoring, health and assistive application,
communication, etc.).
User Information is collected and analysed in a proper User Analysis process,
in order to extract user needs and preferences. Preference data can be gathered
using interviews, workshops, surveys, site visits, artefact analysis, focus groups,
observational studies, and contextual inquiry. The analysis of user needs should
68 Ceccacci et al.
take into consideration the fact that elderly people present a very diverse range of
abilities. This diversity of abilities exists not only within groups, but also among
individuals. Therefore, it is impossible to draw up a simple profile, or to identify a
single stereotypical user. In this context, the International Classification of
Functioning, Disability and Health (ICF) is a valuable tool for understanding the
abilities of a user, depending on his/her characteristics.
Functional
Modelling
Figure 7.1. The proposed UCD approach for inclusive design
The adoption of a Functional Modelling approach allows the system/environment

functions and flows to be structured. In this way it is possible to represent the design
context in terms of actions to be supported and mutually correlated. It is fundamental
to analyse how a particular user can act in an environment according to their
capabilities. For example in a typical kitchen it is possible to perform a particular set
of macro-tasks, such as preparing meals, eating, drinking, doing the housework, etc.
Each macro-task consists of a number of actions which require specific physical and
mental skills to be performed. Once the functional model of the environment has
been built, it is possible to deduce how the lack of special skills impacts on the
performance of each task, and hence to define the characteristics of a clear set of
appropriate facilitators to be provided for a barrier-free kitchen.
Prototyping activity is fundamental to communicate and assess the design intent.
As mentioned above, the creation of high fidelity prototypes represents a cost that
increases if alternative design solutions need to be evaluated. Moreover, at the
beginning of the design stages it is very difficult to build a prototype which is
similar to the final product in terms of implemented functions, adopted materials,
aesthetic features, interaction modes, etc. because all design choices have not been
made yet. In parallel, the involvement of end-users during conceptual design is
fundamental to address subsequent decision-making. As a consequence, thanks to
the advantages of novel AR/VR technologies, high-fidelity prototypes can be
replaced with virtual ones. The problem lies in choosing the technology according
to particular user characteristics. In fact, the experienced quality of a VR
environment, in terms of interaction, immersion and navigation, generally depends
on the human mental and physical ability.
The evaluation of product usability is fundamental for the creation of user-

appropriate products. It allows designers not only to pinpoint design errors, but
also to measure the quality perceived by the user and the resulting product
experience, so that the design changes can be better finalised. This is carried out
through usability tests, which support the collection of detailed information on the
product experience and on the way consumers use it. To fully evaluate product
usability it is necessary to develop an experimental protocol, which allows the
correlation of the user response with the specific product features by considering
emotional, affective and cognitive aspects of the user-product interaction.
7.4 The e-Kitchen Case Study and Experimental

Results
A preliminary study was conducted to assess how much TAR technology can be
useful in involving the elderly in the early usability evaluation of the centralised
control panel of the e-kitchen model (called smart kitchen user interface). The
structure of the e-kitchen system is represented in Figure 7.2. The control panel
manages all household appliances, receives suggestions and feedback from them,
monitors consumption, etc.
TAR usability assessment is carried out through the following steps:
1. definition of an experimental protocol to assess the usability of different
prototyping environments;
2. construction of a traditional high-fidelity prototype and development of
TAR prototype operating as the final system and enabling the user to
interact with the physical elements to trigger an emotional and affective
response;
3. application of the experimental protocol in order to compare user
performance in two different environments (real and virtual).
The proposed experimental protocol aims to measure efficiency, effectiveness
and satisfaction in use provided by logical and physical elements of the interface,
as prescribed in ISO 9241-11 (1998). The protocol is general enough to be able to
be applied in different interactive devices and different virtual reality
environments.
Efficiency was assessed as the completion time required to complete a task.
Effectiveness was evaluated by error occurrence, meaning the number of mistakes
during task completion such as incorrect action sequence, repeated actions or
commands and wrong actions (ISO/IEC TR 9126-4, 2004). This makes it clear if
the product interface correctly guides users toward task completion. Finally, some
metrics have been defined to evaluate satisfaction in use: pleasure in use, absence
of monotony, accessibility and ease of use (UNI EN ISO 10075-2, 2002; Kuijt-
Evers et al., 2004; Park and Han, 2010). Pleasure in use measures perceived
pleasure and satisfaction during use. It can be defined as a mental state depending
on the quality of user-product interaction. Absence of monotony measures how
well the system avoids excessive mental workload and repetitive actions.
70 Ceccacci et al.
Accessibility measures intelligibility and adequacy of the information provided,

and so it is closely connected with the understanding of product behaviours,
functionalities and modes of use. Ease of use measures the simplicity of using the
product interface and the perceived sense of friendliness.
Figure 7.2. The e-Kitchen model
The adopted TAR prototype is characterised by a tangible interaction in an AR

environment and by a functional simulation of the GUI behaviour. The tangible
interaction is achieved by using a physical prototype realised by low cost Rapid
Prototyping techniques. A real scale projection of the virtual prototype is
overlapped with the physical interface to simulate the aesthetic appearance (i.e.
surface finishing, shape) and the visualisation display that changes according to the
user options. Reliable real time interaction is guaranteed by an optical tracking
system (Optitrack by NaturalPoint) which tracks the user’s hand, the user’s point
of view and the product prototype position in the same space. The modelling of the
GUI functional states and the management of state changing in real time is
achieved by 3DVIA Virtools (by Dassault Systems). A portable clip-USB camera
is mounted on see-through glasses (iWear VR920 by Vuzix) to record the scene.
Such a set-up allows users to handle the product prototype and simultaneously see
the virtual interface projected on the physical object.
The high fidelity prototype (HF) is produced by building all the electronic and
mechanical components of the device. It is made of a moulded square containing a
case (PMMA for the front side and TPU for the back), a lion battery, a Bluetooth
module, an electronic board, a LCD 3.5” capacitive APR touch screen, eight
piezoelectric actuators, and a polycarbonate closing ring. The SW system is
developed in C++ on Linux adopting Nokia QT graphical libraries. This set-up
provides a powerful product representation since the highest level of interaction
and fidelity is achieved. However, it entails high costs and a long time to produce
the whole prototype.
Figure 7.3. On the left the TAR prototype (A) the RP model with AR and Optitrack IR
marker and (B) the augmented image of interface. On the right the HF prototype.
Involved end-user composition was defined by adopting a subjective sampling

method based on the field studies previously conducted by industrial partners on
age and gender of the target market. A total of 16 users were involved in the testing
sessions, eight users for each experimental set-up: this number is sufficient for a
qualitative assessment, because it permits the detection of more than 80% of all
usability problems in a product, as supported by different studies (Virzi, 1992;
Nielsen and Landauer, 1993; Nielsen, 2000). The groups were matched in gender
(for each group 75% were women and 25% were men), age distribution (75% 60-
65 years old, 25% were 66-70) and technological expertise with respect to the
different set-ups.
Tests were carried out in parallel groups, as participants were statistically
matched. Each group had to evaluate a product prototype produced according to
the respective test set-up methods.
To evaluate metrics of satisfaction, users were asked to express their personal
judgment for each protocol metric (5-point scale). To evaluate efficiency and
effectiveness, users were asked to perform five different tasks on the product
interface which were defined so as to evaluate the quality of the interaction with
every product function according to the above metrics of measurement. The joint
use of VIA and interviews assured an accurate analysis of user response.
A One-Way ANOVA (alpha level of 0.1) has been calculated in order to
compare experimental data (task completion time, number of errors and
satisfaction in use) in order to analyse differences between TAR and HF set-ups.
On analysing the results, the adopted protocol proved to be appropriate to evaluate
the GUI and the TAR prototype has shown it can be useful to investigate both
cognitive and physical ergonomics due to its dual nature (i.e. virtual interactive
graphic interface and physical interaction with the product body). In particular it is
possible to infer that the technological set-up does not affect task errors and
satisfaction in use. However, time performance is significantly different (Table
7.1).
Results point out the effectiveness of adopting a tangible augmented prototype
to create interactive prototypes for the usability testing.
To prove the effectiveness of TAR techniques to carry out usability assessment
in place of high fidelity prototypes an additional evaluation is imperative. It
consists of analysing the effort required to virtually prototype alternative solutions
and to set the proper experimental environment (Table 7.2). Table 7.2 provides a
time estimation for building the two tested prototypes. The assessment is carried
72 Ceccacci et al.
out considering five main activities in prototyping as listed in Table 7.2. A total
time comparison shows that TAR sensibly reduces the time for prototype
development and testing. While previous results show the reliability of usability
assessment on virtual prototypes instead of high fidelity ones, Table 7.2
demonstrates that designers do not require extra work to build them. Both
achievements are fundamental to increase the use of these technologies in industry.
Table 7.1. One-way ANOVA results for satisfaction, task completion time and errors
However, although this result is crucial to demonstrate the importance of a

UCD approach based on TAR for industrial applications, it is not enough to show
that it is suited to test usability for elderly and disabled people. Although aged
people were chosen to carry out the testing sessions, their response in terms of
usability of the adopted technologies has not been structurally assessed. The
adopted protocol actually aims to evaluate the usability of the control panel
independently from the technique used to realise it. A proper post-hoc
questionnaire should be designed to focus on the user experience of the deployed
prototypes. A preliminary analysis of users’ observations outlines that:
1. they are quite at ease with the AR glasses, which are similar to those they
wear in everyday lives, but they show problems in vision if they need their
own glasses to focus on the displayed images;
2. they have problems in selecting the virtual icon by positioning their finger
in the right place to allow the panel to work properly. This problem is
particularly evident in the presence of slight hand tremors;
3. they have some difficulties in identifying the right distance to position the
physical prototype in respect to their viewpoint filtered by the AR glasses.
Table 7.2. Design process performance resulting from the different prototyping techniques
(evaluation in time)
ACTIVITIES TAR prototype High-fidelity Prototype
1. Modelling of the design alternatives 4 days 10 days
CAD modelling 2 days 4 days
GUI modelling 2 days 6 days
2. Prototyping 5 days 20 days
File import/export 1 day 1 day
Setting (machine, rendering scene, 3 days -
etc.) 1 days 18 days
Prototype generation - 1 day
Finishing
3. Experimental design 0.5 hour -
4. Pre-testing (users’ training, etc.) 1 hour -
5. Testing 3 hours 2.5 hours
TOTAL TIME 9 days 4.5 hours 30 days 2.5 hours
7.5 Conclusions
The use of VR/AR technologies can successfully support the design and evaluation
of Ambient Assisted Living environments. They can facilitate the assessment of user
interfaces from both a cognitive and a physical point of view. In this context an
Inclusive design approach has been defined in the scenario of a funded National
project called e-kitchen. A part of this approach addressed to the VR/AR technology
choice to virtually test different software user interfaces. In this case the technology
assessment has to be properly correlated to a specific class of users (aged people).
Experimentation highlights that Tangible AR is the best system to allow subjects to
experience the interaction. Future work will be focused on a deep analysis of the
usability of the deployed TAR technologies with specific attention elderly people and
people with disabilities.
7.6 References
VAALID (2011) Accessibility and usability validation framework for AAL interaction design
process. Available at: http://www.vaalid-project.org/ (Accessed 4 November 2011)
Alankus G, Lazar A, May M, Kelleher C (2010) Towards customizable games for stroke
rehabilitation. In: Proceedings of the 28th ACM Conference on Human Factors in Computing
Systems (CHI 2010), Atlanta, GA, US
Bordegoni M, Cugini U, Caruso G, Polistina S (2009) Mixed prototyping for product assessment:
A reference framework. International Journal on Interactive Design and Manufacturing, 3(3):
177-187
Flores E, Tobon G, Cavallaro E, Cavallaro FI, Perry JC, Keller T (2008) Improving patient
motivation in game development for motor deficit rehabilitation. In: Proceedings of the
International Conference on Advances in Computer Entertainment Technology (ACE 2008),
Yokohama, Japan
Hall R (2001) Prototyping for usability of new technology. International Journal of Human-
Computer Studies, 55(4): 485-501
74 Ceccacci et al.
ISO 13407 (1999) Human-centred design processes for interactive systems. International
Organization for Standardization. Available at: http://www.iso.org/iso/catalogue_detail.
htm?csnumber=21197 (Accessed 4 November 2011)
ISO 9241-11 (1998) Ergonomic requirements for office work with visual display terminals
(VDTs) - Part 11: Guidance on usability. International Organization for Standardization.
Available at: http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnum
ber=16883 (Accessed 4 November 2011)
ISO/IEC TR 9126-4 (2004) Software engineering - product quality- Part 4: Quality in use metrics.
Available at: http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber
=39752 (Accessed 4 November 2011)
Kim S, Dey AK (2009) Simulated augmented reality windshield display as a cognitive mapping
aid for elder driver navigation. In: Proceedings of the 27th ACM Conference on Human
Factors in Computing Systems (CHI 2009), Boston, MA, US
Kuijt-Evers LFM, Groenesteijn L, De Looze MP, Vink P (2004) Identifying factors of comfort in
using hand tools. Applied Ergonomics, 35(5): 453-458
Kuutti K, Battarbee K, Sade S, Mattelmaki T, Keinonen T, Teirikko T et al. (2001) Virtual
prototypes in usability testing. In: Proceedings of the 34th Hawaii International Conference on
System Sciences (HICSS 2001), Maui, Hawaii, US
Mengoni M, Peruzzini, M, (2010) Usability assessment method to improve interaction design:
How to get it. In: Fisher X, Coutellier D (eds.) Research in Interactive Design, Springer-Verlag,
Paris, France
Milgram P, Takemura H, Utsumi A, Kishino F (1994) Augmented reality: A class of displays on
the reality-virtuality continuum. In: Proceedings of SPIE, Telemanipulator and Telepresence
Technologies.
Nielsen J (2000) Why you only need to test with 5 users. Alertbox. Available at:
http://www.useit.com/alertbox/20000319.html (Accessed 4 November 2011)
Nielsen J, Landauer TK (1993) A mathematical model of the finding of usability problems. In:
Proceedings of the ACM INTERCHI’93 Conference, Amsterdam, The Netherlands
Park H, Moon HC, Lee JY (2008) Tangible augmented prototyping of digital handheld products.
Computers in Industry, 60(2): 114-125
Park YS, Han HS (2010) Touch key design for one-handed thumb interaction with a mobile
phone: Effects of touch key size and touch key location. International Journal of Industrial
Ergonomics, 40(1): 68-76
PERSONA (2007) Perceptive spaces promoting independent aging. Available at: http://www.aal-
persona.org/index.html (Accessed 4 November 2011)
Pittarello F, De Faveri A (2006) Improving access of elderly people to real environments: a
semantic based approach. In: Proceedings of the Working Conference on Advanced Visual
Interfaces (AVI 2006), Venezia, Italy
Sonderegger A, Sauer J (2010) The influence of design aesthetics in usability testing: Effects on
user performance and perceived usability. Applied Ergonomics, 41(3): 403-410
UNI EN ISO 10075-2 (2002) Ergonomic principles related to mental workload - design
principles. Italian Organization for Standardization. Available at: http://www.uni.com/en/
(Accessed 4 November 2011)
UTOPIA (2011) Usable technologies for older people: Inclusive and appropriate. Available at:
http://www.dcs.gla.ac.uk/utopia/default.html (Accessed 4 November 2011)
VERITAS (2011) Virtual and augmented environments and realistic user interactions to achieve
embedded accessibility designs. Available at: http://veritas-project.eu/ (Accessed 4 November
2011)
Virzi RA (1992) Refining the test phase of usability evaluation: how many subjects is enough?
Human Factors, 34(4): 457-468
Wilson JR, D’Cruz M (2006) Virtual and interactive environments for work of the future.
International Journal of Human-Computer Studies, 64(3): 158-169
Zhou Z, Cheok AD, Yang X, Qiu Y (2004) An experimental study on the role of 3D sound in
augmented reality environment. Interacting with Computers, 16(6): 1043-1068
Chapter 8
Development and Evaluation of Sonified

Weather Maps for Blind Users
R. Weir, B. Sizemore, H. Henderson, S. Chakraborty
and J. Lazar
8.1 Introduction
In recent years there have been significant advances in developing websites that are
accessible for individuals with disabilities. In particular, the Web Content
Accessibility Guidelines (WCAG) from the Web Accessibility Initiative
(http://www.w3.org/wai) provide clear standards for developing accessible web
based content, and have been the foundation for legal guidelines developed in
many countries (Meiselwitz et al., 2010). However, a challenging aspect of
accessible web design is developing equivalents of data visualisations for blind
users, since data visualisations are often used to allow quicker comprehension of
large and complex data sets (Fritz and Barner, 1999). In this research we
investigate the potential use of sonification for presenting weather data to blind
users. At this juncture we would like to emphasise that that the term “blind users”
has different meanings depending on context and country. For instance, in the US,
“blind users” refers to anyone with any type of visual loss, whereas in the UK,
“visually impaired” is often used to describe people with low vision, and “blind”
is used to describe someone with no useful residual vision.
Maps represent a very common visualisation approach on web pages. Figure
8.1 (adapted from www.recovery.gov) shows a typical example of map based data
visualisation. This map of demographic data allows the user to quickly reach a
conclusion that southwestern states in the USA have a higher percentage of
residents under 18 through an interpretation of the colour coding in the map.
Equivalents of visualisations for blind users tend to be simple tabular
representations of the same data (Plaisant, 2004). While this is technically
equivalent, it does not provide an “overview of data” that is the key first step of
usage of information visualisations.

76 Weir et al.
Figure 8.1. Map based visualisation of population data
Previous research has addressed this problem through the development of sonified
maps, which used non-textual audio output to allow users to comprehend detailed
as well as trend data rendered within maps (Zhao et al., 2008; Walker and Mauney,
2010). In this investigation, we build on existing research to investigate the
potential use of sonification to represent weather map data for blind users. The rest
of the article is organised as follows. First we describe existing research and
applications related to sonified maps. Next we describe our methodological
approach in developing and evaluating a sonified solution for providing accessible
weather map based data for blind users. Finally, we discuss the implications and
future research directions.
8.2 Literature Review

An important implication of developing websites following accessibility guidelines
such as WCAG is that they will work for most users with perceptual and/or motor
impairments, including blind users. However, information visualisations on web
pages continue to be inaccessible. Visualisations are inherently directed at sighted
users and developing an equivalent rendering for blind users is not a trivial
problem. One option is to use tactile printed media (such as a raised bar chart
printout). However research (e.g. Wall and Brewster, 2006) suggests that these
have limitations both in terms of flexibility of option and extent of usage. Effective
visualisations need to incorporate a capability to drill-down to specific data values
in addition to providing an overview of data (Shneiderman and Plaisant, 2010).
Previous research has examined the potential use of sonification of maps for blind
users, utilising non-textual sound (Zhao et al., 2008; Pauletto and Hunt, 2009).
Two existing sonification tools are iSonic (http://www.cs.umd.edu/hcil/audiomap/)
and Earth+ (http://prime.jsc.nasa.gov/earthplus/). The iSonic tool was originally created
as a graduate project at the University of Maryland and allows blind users to hear
population trends and patterns on a map of the United States. At a very basic level
the application sound pitch to provide the user overview or trend level information
about population within a geographical region. iSonic also provides alternative
Development and Evaluation of Sonified Weather Maps for Blind Users 77
map views of data that include a broader regional view and also a view down to the
state level and associated counties. Earth+ (developed by NASA in 2005) is
another tool that was evaluated. Earth+ provides accessible map interpretation for
blind users. Unfortunately, the Earth + application is limited in its functionality.
The application allows a user to load an image and provides a piano note with a
pitch unique to the colour of the image based on the position of the cursor.
Therefore, we chose to focus our work on using iSonic, and received permission
from the University of Maryland to continue building onto the iSonic application.
8.3 Requirements Gathering

To modify an existing tool for blind users to access new types of data content
(weather data), we needed to understand how blind users access weather
information. We were therefore motivated in our requirements gathering approach
toward answering questions about the different kinds of weather information
considered important by blind users and how it needed to be rendered, resources
available to obtain the requisite weather related data feeds, and the most
appropriate interface technology that could be used.
Our data collection was greatly facilitated by our access to a diverse set of users
within the blindness community. There were four interviews and six survey
responses. Our four interviews took place with individuals who had 1) expertise
about the domain of interest (meteorology), 2) a high level of motivation in helping
develop and use sonified weather maps, 3) expertise on contextually appropriate
assistive technology, and also with one individual who was a meteorologist but
was not blind. In addition, there were also 6 responses to a survey, from typical
users who had interest in gaining weather related information but moderate to low
technological and domain related expertise. Space constraints prevent us from
providing specific details about the interviews and the survey. Users were either
interviewed or surveyed based on their availability, as is typical for human-
computer interaction research involving people with disabilities (Lazar et al.,
2010). We employed interviews when face time was possible, and surveys when
face time was not an option.
The interviews with the expert blind users and the survey responses of the non-
expert blind users provided valuable input for the development of the prototype
weather map. The data from the above user interaction provided requirements for
the application related to the nature of weather related information expected by
blind users, the most useful representations of such information, as well as modes
of interaction that would be considered helpful. In addition to the interactions with
blind users, the data collection also involved an interview with a meteorologist
from the National Oceanic and Atmospheric Administration (NOAA) who is not
blind. This interview provided valuable directions with regards to obtaining a
continuous live weather data feed for the application from NOAA web sites.
After the collection and analysis of interview and survey data, a series of user
scenarios were created to obtain a clearer understanding of what the proposed
application could be expected to do. These scenarios were created to explicitly map
78 Weir et al.
out how a user would go through the program for related tasks, what these tasks
would show, and what types of outputs they would have. Special care was taken to
explicitly include features and capabilities identified in the user interviews and
surveys while developing the user scenarios. The scenarios proved to be helpful in
narrowing down and creating a focus for what the application should do and how.
Based on the requirements gathering and the user scenarios, it was possible to
identify an initial set of specifications for the sonified weather map application.
These specifications are described as follows:
• provide weather information related to a) current and future temperature b)
precipitation and c) wind speed;
• allow the users to obtain discrete weather related information for a
geographical location (e.g. temperature in a particular city);
• allow the users to obtain trends with regards to weather related information
for a geographical region (e.g. change of temperature within a state);
• provide the users with a sense of the physical geographical reference
within the context of the map;
• allow the user to choose a) the nature of weather information and b) the
level of detail for the weather information;
• provide the user with means for multimodal interaction to enable
perception of information through multiple senses;
• be easy to use for users without access to sophisticated assistive
technology;
• be easy to use for users with minimal experience with assistive technology.
The above list of specifications provided the basis for the design of the sonified
weather map application. In addition a decision was taken to develop the initial
prototype only for the state of Maryland and the 24 counties within Maryland
(Baltimore City is not in a separate county, and demographically is usually counted
as the 24thcounty of Maryland). The specifics of the design are described in the
next section.
8.4 Design of Sonified Weather Map

8.4.1 System Architecture for Retrieving Weather Data
The accessible weather map was built by modifying the iSonic application. The
iSonic application in its current version works with static population data and
requires the source data to be in a CSV file format. However, weather data is more
dynamic, changing often at an hourly rate. Design modifications were therefore
required to accommodate the dynamic data.
The source for the weather data was the National Oceanic and Atmospheric
Administration (NOAA) website. This website provides real-time weather data
formatted for easy retrieval. The data querying was developed using Microsoft
Excel macros that allow retrieval of external data from tables located on HTML
pages. The weather data for the different counties within Maryland was retrieved
through the manipulation of the longitude and latitude values in the data retrieval
queries. This allowed the importing of weather data for multiple points in
Maryland into the excel sheet. The particular data points were developed through
the compilation of a list of cities that were completely within the boundaries of
each county, usually the county seat. The assumption was that these cities would
be representative of the weather within the county in most instances, since most of
these counties are not geographically large.
A second Excel sheet was created and formatted appropriately to make it
readable by the iSonic application. The cells of the two excel sheets were linked so
that the second excel sheet could automatically pull data from the first excel sheet
containing imported weather data from NOAA’s web page. A small java program
was developed that that allowed automatic storing of the data from the second
excel sheet as a CSV file. In addition VBA macros were created for both excel
sheets that initiate a refresh, save, close window sequence. Finally a batch file was
created so that, the weather data is automatically updated and converted into a file
readable by the application. A system architecture schematic for the data retrieval
process is shown in Figure 8.2.
Supplies Supplies
NOAA Web Server Master Data Excel Sheet
Converts to
Is read by
Sonic Map Interface
CSV File
Figure 8.2. System architecture schematic
8.4.2 User Interaction with the iSonic Weather Map

The prototype iSonic weather map application was developed by overlaying
weather data on the basic iSonic Application. The application design was therefore
constrained to use the existing mapping between data and auditory feedback in this
platform. An associated risk of this design decision is the possibility that the
mapping would not effectively translate to a different domain (weather data).
While this represented a design dilemma, a decision was taken to accept this as a
constraint for this exploratory research, and also to use this as an opportunity to
test the usefulness of the existing mapping.
The prototype weather map was developed to provide the user with three
different types of weather information – temperature, wind speed and percentage
chance of precipitation (see Figure 8.3). These types of weather related information
are presented to the user separately to reduce information clutter. The user has the
ability to switch between three sonified maps.
80 Weir et al.
The primary interaction mechanism in the iSonic interface is keyboard-based.

For the study, an additional interaction strategy was adopted, involving a touch
screen with a tactile map overlay to traverse through the iSonic weather maps. A
KeyTec Magictouch touch screen was used to evaluate the potential for touch-
screen based interaction with the application. The initial plan was to use a tactile
map of Maryland obtained using the Braille/tactile printing facilities at the
National Federation of the Blind. However, this tactile map did not calibrate well
to the scale expected by the touch screen. A work around was achieved, by tracing
the map of Maryland from the iSonic interface and then creating tactile contours of
the map boundaries using needles, on a sheet of paper that was thick enough not to
rip, but thin enough that the touchscreen could still sense human touch. This tactile
map overlaid on the touch screen calibrated well to the iSonic application.
The iSonic interface provides two information presentation choices (or data
views) to the user. The first is the default map-based representation. The second is
a tabular data view where the first column represents the geographical elements of
the data (counties of Maryland in this instance) and the other columns represent the
specific data domains being presented (e.g. temperature, wind speed and
precipitation, in this case). The user can switch between the two data views using
the TAB key. The tabular data view allows the user to toggle between the three
different weather maps (for temperature, wind speed, and precipitation).
iSonic provides the user with two different choices of navigation through the
map - absolute and relative. The absolute navigation allows the user to sweep
through (from left to right and up to down) and provides an overview of map
values. The interface provides a percussion sound at the end of each row and bell
sound at the end of the sweep. For relative navigation the user can press the four
arrow keys to navigate up, down, left and right in the map commencing from the
left top corner of the map. A chirping sound alerts the user when the navigation
takes him/her outside the boundaries of the map. The map is always traversed in
units defined by the geographical boundaries of the Maryland counties. As the user
traverses through the state, they are given auditory feedback about the weather
related data for that county. The user has the choice of receiving this feedback with
different degrees of detail or information level. The lowest level information is
provided by a tonal sound. The pitch of this sound is used to represent the value of
the data, a higher pitch being associated with higher value of data. The higher
level of information augments this tonal sound with spoken feedback, such as the
name of the county and the temperature in the county. For this study the users were
also provided with the means of navigating using the touch screen and the tactile
map overlay. With this alternative, the authors were able to select a particular
county by tracing its contours in the tactile map and then tapping on it to get
weather related auditory feedback for that county.
Figure 8.3. Prototype map interface
8.5 Usability Evaluation

An important aspect of developing any application for users with disabilities is to
do a usability evaluation. Furthermore, the previous research (Zhao et al., 2008) on
iSonic only evaluated the software using the keyboard, so we felt that evaluating
using the touchscreen functionality would provide useful feedback for both
developers and researchers. We also wanted to learn more about the effectiveness
of just a touchscreen as compared to a touchscreen with tactile overlay. As this
research project had no funding, and the volunteer participants did not have much
time available, it was not possible to conduct the usability testing in stages over
months and determine the effect over time of getting experience with the
application. So this was a shorter, more formative usability evaluation, rather than
a summative evaluation, which also focused on qualitative, rather than quantitative
evaluation. There were a large list of potential tasks, such as naming the five
counties with the lowest temperature, comparing the temperatures in two counties
that bordered a third county, assessing trends in temperature (e.g. “does the
temperature rise or fall as you go from western Maryland to central Maryland?”),
where in the state are the windiest counties located, and what portion of the state
has the highest precipitation. Due to the large number of potential tasks and the
limited time for usability evaluation, the participants were mostly asked to perform
different tasks. We planned for testing to take place with five blind individuals, at
the International Braille and Technology Centre in Baltimore, Maryland. These
individuals were recruited through the National Federation of the Blind, and had
specifically expressed interest in weather maps, sonification, or science education.
Two of the participants in the usability evaluation were the same people who took
part in the interviews discussed earlier (the Blind user with expertise in
meteorology, and the Blind user interested in developing accessible weather maps).
Our 5 blind individuals had an average age of 45.6 (range 28-68), and there
were 4 males, 1 female, all of whom were screen reader users who are unable to
use screen magnification. The participants had used computers an average of 30.8
years, and used screen reading software an average of 22 years. Clearly, the
participants had a high level of computing experience: however, it is expected that
any potential users of sonification software for weather maps would have a high
82 Weir et al.
level of experience. Of the five participants, only one had previously used any type
of sonification software (for computer gaming).
As the software was being set up before the usability evaluation, there turned
out to be technical problems. Due to some security patches recently installed at the
International Braille and Technology Centre computers, the sonification tones
would not work on any of the computers there. The researchers made a last minute
decision to use the Macbook laptop (booted in Windows mode) that they brought
with them to take notes, to have the participants evaluate instead the keyboard
version (including the sonification tones), and then used a PC in the International
Braille and Technology Centre (without the sonification tones) to evaluate the
effectiveness of the touch screen and the touch screen with tactile overlay. This
was not ideal, but since the usability evaluation was formative and qualitative, it
allowed for a lot of feedback from users and discussion. The researchers explained
to users how the software worked and demonstrated it using keyboard and
touchscreen, gave the participants a few minutes to play around with it, and then
asked the participants to attempt some tasks. In general, all five participants liked
the application, and were able to figure out how to use it, to successfully complete
a few tasks, within a few minutes of first using the application.
User 1 spent several minutes trying to get a feel for the program, and initially
said that they liked the keyboard application but they wanted to be able to feel the
edges of a tactile map. When using the touchscreen, the user perceived the
touchscreen as being a little bit jumpy and stated that they didn’t like a touchscreen
without a tactile overlay, preferably on the touchscreen, as compared to the
keyboard. User 1 also noted that they listened more to the speech than to the tones.
One interesting challenge is that user 1 assumed that the touchscreen was a multi-
touch screen, which it was not.
User 2 did not like using the keyboard to navigate around the Maryland state
map, but quickly got comfortable with the application and noted points like
“central Maryland is definitely hotter” and “there’s a weather front somewhere
here.” User 2 said that they didn’t like the tones, because “I’m not musically
inclined, so I like numbers, not sounds, it’s my learning style.” User 2 also
expressed a strong preference for the tactile overlay on the touchscreen. User 2 also
felt there could be more clarity in the feedback provided for navigating off the map
and suggested, “You have entered Virginia” or something similar. User 2 felt that
the irregular shapes of the counties made it difficult to navigate on the keyboard
but suggested that the software would be good for geography lessons.
User 3 also stated that they listened more to the speech data than the tones, but
that they could understand there was a difference in the tones, and they wondered
if headsets would be more useful to perceive stereophonic sounds. When using the
touchscreen, the user stated that they preferred this method, as it enabled non-
sequential navigation. User 3 didn’t seem to have a preference as to the
touchscreen with or without the tactile overlay.
User 4 had a good sense of the different tones and their correlations with the
data and understood the trends. Unlike the other users, user 4 seemed to find the
tones to be very useful. Like user 2, they noted that this application would really be
useful for learning the geography of a new area. User 4 wondered why, when you
cross the Chesapeake Bay, a large body of water, it didn’t make a “splash” sound,
instead of a “chirp” sound (which is the current sound for crossing a body of
water). Using the tones, user 4 could immediately determine that the west side of
Maryland had the highest chance of precipitation, and wondered if the researchers
could add “elevation” to the software application, to help them learn more about
the geography of Maryland. When user 4 started using the touchscreen, the
application crashed, and while it rebooted, user 4 had to leave for a work-related
appointment. Therefore, user 4 was the only one who was not able to evaluate the
touchscreen interaction with the application.
User 5 liked being able to hear the trends using the sonified tones, and
immediately picked up important trends, for instance, the chance of rain was higher
in the northern and western parts of the state. User 5 really enjoyed using the
application, and wondered how much data you could present to a user before they
became overwhelmed: over-time comparisons (e.g. checking the data at 7AM and
again at 11AM) might be most useful. User 5 was equally enthusiastic about the
tactile map over the touchscreen, and was able to easily complete tasks using both
approaches, commenting “now it starts to mean something, because now I’m
touching it on the map.” User 5 further noted “Now, I get the information that I
don’t normally get. This is a very different sense than I get from using the Braille
note. This is exactly what I have been looking for!” User 5 further noted “I’ve
always had to calculate the weather trends in my mind, until today!”
In summary, while the usability evaluation was formative and qualitative, there
were some clear trends. The users clearly preferred the tactile map over the
touchscreen, as compared to either the touchscreen alone, or keyboard alone. While
some users found the sonification tones useful, other users did not. Two users who
had recently moved to Maryland thought that this software application would be
very useful for learning state geography, which was not a stated scenario or
development goal for the project, but could be a potential goal. Suggestions for
improvement included a textual notification when you left a state border (such as
text saying, “You are now in Virginia”), a splash sound instead of a bird chirp to
notify you when you are crossing a body of water, and headsets to get a better
spatial sense of where the sounds are coming from.
8.6 Implications and Discussion

While the current investigation into developing sonified weather maps is at an
early stage, there are some interesting implications for future research. Feedback
from users indicates that visualisation maps present important and commonly used
data representations that remain mostly inaccessible to blind users. This
inaccessibility becomes more pertinent in information contexts with an inherent
spatial component (such as weather data) where data values change rapidly and
map sweeps can lead to important trend analyses. The enthusiasm of our usability
evaluators (during interactions with the touchscreen and the tactile map) at being
able to, for the first time, get an idea of the spatial orientation of the state of
Maryland on the computer, indicates that there are interesting implications of
research into designing and employing accessible maps as a pedagogical tool for
84 Weir et al.
learning not only weather, but also geography. Furthermore, because mobile
phones and tablet devices are increasingly using multi-touch screens, and weather
information is often needed on-the-go, it is important to investigate in the future
how these applications could be used in portable devices such as Droid Phones and
the Apple iPad. Touchscreens can be fully accessible to blind users (by using
speech output, touchscreen gesturing, and Braille overlays to indicate where the
visual keyboard appears), and it seems that a next step for research might be to
make the application more robust, and determine how it might be implemented on
a mobile or tablet device.
8.7 References
Fritz JP, Barner KE (1999) Design of a haptic data visualization system for people with
visual impairments. IEEE Transactions on Rehabilitation Engineering, 7(3): 372-384
Lazar J, Feng J, Hochheiser H (2010) Research methods in human-computer interaction.
John Wiley and Sons, Chichester, UK
Meiselwitz G, Wentz B, Lazar J (2010) Universal usability past, present and future.
Foundations and Trends in Human Computer Interactions, 3(4): 213-333
Pauletto S, Hunt A (2009) Interactive sonification of complex data sonic interaction design.
International Journal of Human-Computer Studies, 67(11): 923-933
Plaisant C (2004) The challenge of information visualization evaluation. In: the
Proceedings of the Working Conference on Advanced visual interfaces, Gallipoli, Italy
Shneiderman B, Plaisant C (2010) Designing the user interface: Strategies for effective
human-computer interaction, 5th edn. Addison-Wesley, Boston, MA, US
Walker BN, Mauney LM (2010) Universal design of auditory graphs: A comparison of
sonification mappings for visually impaired and sighted listeners. ACM Transactions on
Accessible Computing, 2(3): 12
Wall S, Brewster S (2006) Feeling what you hear: Tactile feedback for navigation of audio
graphs. In: Proceedings of CHI 96 Workshop on Human Factors in Computer Systems,
SIGCHI, Montreal, Quebec, Canada
www.recvery.gov. (Accessed on 28 November 2011)
Zhao H, Shneiderman B, Plaisant C, Lazar J (2008) Data sonification for users with visual
impairments: A case study with geo-referenced data. ACM Transactions on Computer
Human Interactions, 15(1): 4
Chapter 9
Achieving Inclusion in Public Spaces: A

Shopping Mall Case Study
Y. Afacan
9.1 Introduction
Designing inclusive spaces can be seen as a response to accommodate diverse
people within the built environment as efficiently, effectively, and satisfactorily as
possible, regardless of health, body size, strength, experience, mobility and/or age.
Although technological innovation has brought many benefits into architecture and
planning, it is still difficult to embed inclusive design into real-world applications.
Reviewing the literature on inclusion in the architectural design context indicates
that an in-depth understanding to the diverse user of matching marketing purposes
is lacking. Consequently, defining the user in the built environment as an ‘average
person’ creates user-unfriendly public spaces. “Design exclusion does not come
about by chance: it comes about through neglect, ignorance and a lack of adequate
data and information” (Cassim et al., 2007). One of the main reasons for that is:
most design practitioners are unable to take inclusive design into account during
the initial phases of the design process, which leads to wrong decisions that can
have a large impact on the overall design success and cost. The second reason is
related to theory-practice inconsistency (Afacan and Erbug, 2009). Although there
are guidelines and accessibility standards, designers have difficulty in utilising this
academic source of information (Gregor et al., 2005). However, Nicolle et al.
(2003) added, “Designers are under a great deal of time pressure: if knowledge is
not presented in a usable format, it will be either discarded or ignored.” Therefore,
although most designers are aware of universal design, problems appear in the
integration of theories and guidelines into design practice (Demirkan, 2007).
Despite the extensive literature on inclusive design, it is not easy to navigate
the mass of data and interpret it into the cultural context. In Turkey, in the last
decade, there has been a rise in the number of elderly and disabled people. It is
traditional in Turkey to give a place of respect to the aging population in public
spaces. However, compared to Europe and the US, there are still problems of
integration of elderly and disabled people into social life because of environmental
barriers, such as lack of ramps, disabled toilets, and inaccessible entrances to

86 Afacan
buildings. According to the Turkey Disability Survey (2002), most disabled people
are still excluded in public spaces because designs do not provide the same
opportunities of use for all users. Although the Turkish government realised the
importance of inclusion within built environments and is developing policies
(Republic of Prime Ministry Administration for Disabled People, 2011), still there
is a need of studies to promote a positive attitude to inclusive design in the public,
to encourage designers to design inclusively, and to make society sensitive and
informed about diverse user needs, capabilities and expectations.
This study is a further step of the previous study by Afacan and Erbug (2009),
which promoted an interdisciplinary heuristic evaluation process for the universal
usability of shopping malls. Afacan and Erbug (2009) highlighted the importance
of working together with various design professions to lead to inclusion in real
applications. According to this study, the lack of empathy for the requirements of
diverse users is one of the three critical issues that make it difficult to integrate the
inclusive design into current design practice. So, this study now delves deeper into
how to include shopping malls from the user perspective. The aim of the study is
not to evaluate the building performance of the case building, rather to focus on
what makes a shopping mall more inclusive and how important each of the
universal design criteria is in defining a mall as a user-friendly public space.
9.2 Inclusion in Public Spaces: Shopping Malls in

the Turkish Context
Achieving inclusion means embracing difference and celebrating human
diversity. Inclusiveness is crucial for design success, business power and socially
responsive societies. Since a public space is defined as a building that must
accommodate everybody regardless of age, ability and size, the architectural
features of those spaces need to adapt to these differences (Grosbois, 2001).
Everyone needs to be part of society through the use of public buildings (Build
for all manual, 2006). “To achieve this, the built environment, everyday objects,
services, culture and information - in short, everything that is designed and made
by people to be used by people - must be accessible, convenient for everyone in
society to use and responsive to evolving human diversity” (EIDD, 2004). In that
respect, focusing on enabling environments, which means featuring physical and
intellectual accessibility and the sustainability of built structures, together with
their impact on work, mobility and leisure within the community (Build for all
Manual, 2006), will help to make the world more inclusive, bringing the
government on board and engaging with users, in terms of expectations and cost
implications.
Public space in the study is exemplified by a shopping mall. Shopping malls are
particularly important for leisure activities in large urban centres, which should
ensure that all people are equally welcome and that all visitors can participate in
facilities that have no design stigmatisation and that enrich their lives and enhance
autonomy and flexibility (Resolution ResAP 3, 2001). Inaccessible public
buildings for leisure activities are obviously holding disabled people back from
Achieving Inclusion in Public Spaces: A Shopping Mall Case Study 87
productive spheres of society (Haque, 2005). Moreover, the changing leisure and
consumption patterns of Turkish people have made shopping malls among the
most important additions to urban life in Turkey (Erkip, 2003). Within ten years
(from 2000 to 2010), the number of shopping malls in Turkey has increased
tremendously due to structural reforms and the introduction of foreign capital.
“Crowding, traffic problems, and lack of pedestrian safety in the city centre served
to create demand for these new areas” (Erkip, 2003). Although these malls
currently provide a modern well-maintained atmosphere, their spatial and social
characteristics still leave much to improve in terms of common activities, social
participation, independence and well-being. As Lebbon and Hewer (2007) say,
researching inclusion in those spaces is divided between three broad communities:
business, design professionals and design education. However, the user is always at
the heart of these three paths. Thus, seeing user issues from a wider perspective
will not only encourage designers to design for inclusion, but also help society to
increase awareness about enabling environments and develop empathy with others.
Therefore, this study does not limit the impact of a shopping mall to the field of
consumption only, but also highlights its importance in supporting inclusion as part
of social and environmental considerations.
9.3 Methodology
9.3.1 Data Collection
This exploratory study on investigating diverse users’ needs, capabilities and
expectations was carried out in a shopping mall in Ankara, Turkey. The selected
building is one of the biggest of six shopping centres in Ankara. It has an indoor
area of 12.70m2 over nine storeys (five storeys for leisure facilities and four storeys
for car parking) and was built in 2007. It also includes a hypermarket, 195 shops,
nine cinemas, cafes, food court, an entertainment centre and offices.
A survey instrument with a comprehensive list of 110 items was developed to
gather data. It is based on a structured questionnaire format with close-ended
questions. As in Afacan and Erbug’s study (2009), the questions in the survey
instrument were grouped under five categories with reference to the seven
universal design principles. Based on Danford and Tauke’s (2001) definitions these
five essential design elements of a universal city, which should be considered when
applying the seven principles of universal design in built environments, are as
follows:
1. circulation systems: ramps, elevators, escalators, hallways, and corridors;
2. entering and exiting: identifying and approaching the entrance and exit and
manoeuvring through them;
3. wayfinding: paths/circulation, markers, nodes, edges, and zones/districts;
and graphical wayfinding: text, pictogram, map, photograph, and diagrams;
4. obtaining product/services: service desks, waiting areas, and shops;
5. public amenities: public telephones, restrooms (toilets), and seating units.
88 Afacan
9.3.2 Procedure
A total of 120 randomly selected users participated in the survey, including 40
adults (between ages 25-55), 40 elderly (between ages 56-85) and 40 adult with
impairments including 20 physically impaired adults (between ages 28 and 51)
using wheelchairs (n = 13), prostheses (n = 7) as mobility aids and 20 visually
impaired adults (between ages 30 and 59) having total loss of sight (n = 7) and
mild loss of sight (n = 13). The data were collected during face-to-face surveys
with all the participants in a café of the mall. At the beginning, a brief summary of
the procedure and the aim of the study was explained. In the survey, participants
were asked to rate their importance level for each item on a scale of 1-5, (1 being
the least important and 5 the most important) and to mark the appropriate boxes to
identify how important each features is in spending time satisfactorily and
comfortably in a inclusive public space. The items that may not have been clear to
participants were explained as part of the questionnaire. Further information was
obtained through an unstructured interview. Further, to avoid any biases,
participants were not allowed to listen to others while they were being surveyed.
9.4 Results and Discussion

The ratings of the participants on 110 items were analysed with the Statistical
Package for the Social Sciences (SPSS). First, exploratory factor analysis was used
to carry out data analysis. Through the Varimax method, a frequently used rotation
option (Argyrous, 2005), a rotated component matrix was constructed to identify
the number of potentially interpretable factors among the set of correlations within
the data. The matrix indicated the extracted factors with their factor loadings.
9.4.1 Development of Inclusive Public Space Factors

Before carrying out factor analysis, firstly the survey instrument is checked
whether there are any items creating floor and/or ceiling effects. It means that
items at the extreme ends should be eliminated. Since the scale used in the study is
5, items below 1.5 and above 4.5 are regarded as extreme ends. There were no
items at the extreme ends. Secondly, the strength of the correlations among the
survey items was calculated through exploratory factor analysis, which helps
identify common issues and exclude unrelated ones. Pearson product-moment
correlations of the response scores were calculated and a correlation matrix was
constructed, in which all the response items were illustrated in rows and columns
of statistical relationships with a correlation score. Items having a correlation score
lower than 0.30 are avoided for the study, because for a useful statistical approach
a correlation coefficient of 1.00 indicates a perfect association between two
variables (Argyrous, 2005). However, in the study all correlations between item
response scores were greater than 0.30.
The study defined factor loadings in excess of 0.55 as suitable and excluded
factors with factor loading values below 0.55. Total variance of factors was
calculated. “Total variance shows all the factors extracted from the analysis along
with their eigenvalues, the percentage of variance attributable to each factor, and
the cumulative variance of the factor” (Mieczakowski et al., 2010). So, factor
analysis resulted in a five-factor solution that accounted for 54.646% of the total
variance, 110 items had 54.646% variances in common, so they correlated highly
with five common themes; each theme was considered to be a factor (Table 9.1.).
Table 9.1. Total variance explained

Factor Scale Eigenvalue Variance Cumulative
1 Circulation consistency with 8.041 21.385 14.078

user expectations
2 Ease of access to services 3.698 11.239 22.340
3 Multiple path of travel 3.054 7.361 32.124
4 Comfortable use of services 1.990 6.123 42.125
5 Appropriate use of tactile and 1.537 5.038 54.646
audio-visual design features
The inclusive meanings assigned to the five factors are explained below:
1. Factor 1, ‘circulation consistency with user expectations’ is defined by
equitable and simple use of the stairs, moving ramps, elevators and escalators. The
appropriate uses of the tactile, aural, visual design features to maximise their
legibility are as important as ease of use of circulation elements. Variables on this
factor also include provision of clear surfaces for effective manoeuvring, which is an
essential design consideration of public spaces for physically disabled people.
2. Factor 2, ‘ease of access to services’, deals with using shops, waiting desks and
other public services with low physical effort and equitably. Walking along
unimpeded should be a consideration for all people. Any level changes can create
barriers for all disabilities and should be avoided or replaced by gentle slopes.
3. Factor 3, ‘multiple path of travel’, is defined by flexibility and simplicity of
circulation, entering/exiting and way finding. Diverse choices of these elements help
create inclusion in public spaces. Variables on this factor also include entering and
exiting with low physical effort.
4. Factor 4, ‘comfortable use of services’, is defined as being use with low
physical effort. Comfort in the public spaces can be achieved with lighting, public
seating and sheltering structures. A calm, welcoming, user-friendly atmosphere of the
shops and urban facilities is required by everyone (Burton and Mitchell, 2006). All
components of the services should be designed to be comfortable and safe to reach.
5. Factor 5, ‘Appropriate use of tactile and audio-visual design features’, is
defined by the provision of perceptible information. Public space should help all
people regardless their ability to understand where they are and guide them the way
they need to go. Legible spaces with clear signs and tactile surfaces are easy to
navigate.
90 Afacan
Most of the participants had lots of interesting and useful ideas and comments on
how to improve a shopping mall and public space. Table 9.2. lists commonly made
suggestions.
Table 9.2. Reported suggestions for improvement of a public space

Design Elements Suggestion for improvement
Circulation More elevators
Smooth ramps
Wider pedestrian walks
Adequate clear space in circulation elements
Entering and exiting Ease of access to entrances/exits
Adequate dimensions
Clear path to/from the site
Close to car park
Wayfinding Understandable signs
Ease of navigation
Simple layouts
Obtaining product/services More and comfortable waiting areas
Wide passages in shops
Ease of reach to all products
Knee spaces at desks
Public amenities More disabled toilets
Audio-visual features in ATMs
Ease of reach to public telephones
General Audio-visual design features
Textured floor material
No obstructions on pavements
Enough car parking
Knee clearances
9.4.2 Correlation Differences in Diverse User Groups

The study also utilised ANOVA analyses with Bonferroni post-hoc comparisons on
each factor scale score and calculated the F-ratio in order to analyse whether the
scale means of the user groups were significantly different from each other. The
study found statistically significant differences between the user groups in factors
1, 3 and 5. Both the physical and visual impairments of users can affect the design
and usage of a public space. For factor 1 and 3, the mean values indicated that most
of the adults found simple and intuitive use of means of circulation moderately
important, whereas physically impaired and elderly participants consider using
circulation with low physical effort more important. Most of the physically
impaired participants want to enter/exit easily and see multiple choices for entering
and exiting. For visually impaired, it was difficult to use elevators without audio-
visual systems, so they found legibility more important. During the unstructured
interviews regarding circulation, the elderly participants stated that they have a fear
of falling and getting lost while they are spending time in a public space. For factor
5, most of the visually impaired participants emphasised the importance of having
perceptible information for the use of public space elements (toilets, public phones,
and doors). Since most of the services do not make use of a variety of techniques,
such as colour-contrasts, Braille markings, large-print readouts, 16 of 20
participants with visual limitations had difficulties in knowing where and how to
use what. However, the others (adults, elderly and physically impaired) considered
safety features and warning of hazards more important. Furthermore, all
participants regardless of their ability or disability found equitable use of public
amenities very important.
9.5 Conclusions
This study shed light on the needs, capabilities and expectations of diverse user
groups in a shopping mall. The majority of the participants (regardless of their
ability) stated that current real-world applications do not consider diverse user
expectations and public spaces are designed for an average person which leads to
exclusion. The most commonly offered improvements are understandable signs
and ease of navigation. The graphics in signs are small to read for the elderly and
difficult to understand for visually impaired people. Regarding navigation, all
people experience problems because of obstructions and level changes.
The results of the study relate highly to the design principles and
recommendations that have been explained by Burton and Mitchell (2006) for
inclusive urban design. According to Burton and Mitchell, there are six key design
principles; (1) familiarity, (2) legibility, (3) distinctiveness (4) accessibility, (5)
comfort, (6) safety, which make urban life more inclusive, easy and enjoyable for
all members of society. Although these six principles are suggested for ‘streets for
life’, both they and the factors developed in the study emphasise the urgent
necessity of allowing equal access and opportunity regardless of ability and size.
The results of this study also provided an understanding of the importance levels
and attitudes of users towards inclusive environments that maximise quality of life.
The developed factors highlight the significance of a user-friendly public space,
which provides many ways of contact for elderly and disabled people. Since high
quality in public space design is also a key consideration for sustainable
communities both in Turkey and all over the world, equality of access and
opportunity should be achieved to meet inclusion targets and to eliminate the
disabling effects of built environments. However, more analysis should be
conducted in other public spaces and outdoor areas, such as restaurants, cafes,
museums, theatres, libraries and parks. Future research will continue to develop
methods and tools to help designers achieve inclusion in public spaces.
92 Afacan
9.6 References
Afacan Y, Erbug C (2009) Application of heuristic evaluation method by universal design
experts. Applied Ergonomics, 40(4): 731-744
Argyrous G ( 2005) Statistics for research. Sage Publications, London, UK
Build for all Manual (2006) Promoting accessibility for all to the built environment and
public infrastructure. Available at: http://www.build-for-all.net/en/reference/ (Accessed
on 28 October 2011)
Burton E, Mitchell L (2006) Inclusive urban design: Streets for life. Elsevier, Oxford, UK
Cassim J, Coleman R, Clarkson PJ and Dong H (2007) Why inclusive design? In: Coleman
R, Clarkson PJ, Dong H and Cassim J (eds.) Design for inclusivity: A practical guide to
accessible, innovative and user-centred design. Gower Publishing Ltd, Hampshire
Danford GS, Tauke, B (eds.) (2001) Universal design: New York. Mayor's Office for People
with Disabilities, NY, US
Demirkan H (2007) Housing for the aging population. European Review of Aging and
Physical Activity, 4(1): 33-38
Erkip F (2003) The shopping mall as an emergent public space in Turkey. Environment and
Planning, A 35(6): 1073-1093
EIDD (2004) European Institute for Design and Disability. The EIDD Stockholm
Declaration. Stockholm, Sweden
Gregor P, Sloan D, Newell A (2005) Disability and technology: Building barriers or creating
opportunities. In: Zelkowitz M (ed.) Advances in computers. Elsevier, Amsterdam
Grosbois LP (2001) The evolution of design for all in public buildings and transportation in
France. In: Preiser WFE, Ostroff E (eds.) Universal Design Handbook. McGraw-Hill,
MA, US
Haque S (2005) Accessibility for all: Role of architects to make a barrier free environment.
In: Proceedings of UIA Region IV Work Programme ‘Architecture for All’
UIA/ARCASIA Workshop, Istanbul, Turkey
Lebbon C, Hewer S (2007) Where do we find out? In: Coleman R, Clarkson PJ, Dong H and
Cassim J (eds.) Design for inclusivity: A practical guide to accessible, innovative and
user-centred design. Gower Publishing Ltd, Hampshire
Mieczakowski A, Langdon PM, Clarkson PJ (2010) Investigating designers’ cognitive
representations for inclusive interaction between products and users. In: Langdon PM,
Clarkson PJ and Robinson P (eds.) Designing inclusive interactions, Springer-Verlag,
London
Nicolle C, Rundle C, Graupp H (2003) Towards curricula in design for all for information
and communication products, systems and services. In: Proceedings of the International
Conference on Inclusive Design and Communications (INCLUDE 2003), London, UK
Republic of Prime Ministry Administration for Disabled People (2011) Available at:
http://www.ozida.gov.tr/ENG/ (Accessed on 27 October 2011)
Resolution ResAP 3 (2001) Towards full citizenship of persons with disabilities through
inclusive new technologies. Available at: http://www.coe.int/t/e/social_cohesion/soc-
sp/ResAP(2001)3E.pdf (Accessed on 14 July 2011)
Turkey Disability Survey (2002) Available at: http://www.ozida.gov.tr/ENG/ (Accessed on
28 June 2011)
Chapter 10
Visibility Prediction Software: Five Factors

of Contrast Perception for People with
Vision Impairment in the Real World
H. Dalke, A. Corso, G. Conduit and A. Riaz
10.1 Contrast and Vision Impairment

The use of colour contrast in the built environment for people with low vision has
been largely unsupported for architects, access consultants or designers, with little
information available and no easy-to-use tools. Accessible environments assist
everyone including vision impaired people (VIP); yet often people can be disabled
by buildings, not directly by their impairment (Pullin, 2009); in 2002 a total of
0.6% of the world’s population were listed as blind (Harle and McLannahan,
2008). A recent critique of accessibility recommendations showed a lack of
understanding of the five key factors we identified for predicting an object’s
visibility (Dalke, 2011) namely, visual ability (VA) of the observer, contrast, lux
level, dimension of the object and distance away from the observer. These were
established as fundamental for the perception of objects, texts or elements for
VIPs, contrast being one of those five interdependent variables (Dalke et al.,
2010); they are fundamental to the software that has now been developed to predict
object visibility. The research carried out revealed gaps in how to achieve contrast
practically for the professionals who should be more familiar with the process. In
the USA, the ADA Standard for Accessible Design makes reference to contrast but
it is ambiguous and open to interpretation. In the 1991 standard, 70 points of
contrast difference are prescribed for marking warnings on walkways (ADA,
1991). But how to gauge contrast, by calculating the difference between the light
reflectance values (LRV) of two surfaces, is always missing. In the UK there is no
advice on how to check and deliver on contrast for accessible buildings and
products (DDA, 2004; EHRC, 2010).
Previous studies conducted in a laboratory, in simulated real-world scenarios
and on real-world sites (Dalke et al., 2010), indicated that contrast difference is
critical for perception of the world for people, especially those with low vision
(Rogers-Ramachandran and Ramachandran, 1998). Luminance is vital for visual

94 Dalke et al.
accessibility (ANSI/IESNA, 2007); it is not the absolute difference in luminance

that is important, but the relative difference, expressed as contrast (Barten,
1999).
Colour is not that significant for accessibility. It can increase perception if
people have good colour vision, but 8% of the total male population may be
colour vision impaired; people may also be coping with multiple disabilities
(Goldsmith, 1967). There is an infinite variation in the visual capacity, acuity,
and fields of vision of the partially sighted community so perception prediction is
a challenge.
In real-world investigations into what visually impaired people actually see,
and how they use contrast and lighting to navigate, we identified key factors for
perception and defined V4+ mathematically as a boundary for the software by
averaging participants’ results during the tests. Findings from research highlighted:
the disabling effects of glare from white surfaces with black text, rendering them
painful or even impossible to read; the beneficial impact of raising lux levels
further by even just 50 lux; the importance of graduating lighting levels to assist
adaptation to changes from exterior to interior illumination. Not surprisingly the
performance of VIP participants was better in lab tests, without the compounding
variables of visual noise in real-world settings. The final iteration of the software
was derived from experimentation in everyday environments, with VIPs and
released as an app on the iPhone using the five interdependent variable factors; two
of the five are fixed - the VA at 4 and currently the lux at 400. Contrast, dimension
and distance are selectable using sliders on the screen which deliver a result of
VISIBLE or NOT VISIBLE. The app has been tested in real-world scenarios and
in this paper we describe the sequence of these Phases 1, 2, 3 and 4, which
informed the algorithms for software ‘A’, ‘B’, ‘C’ and now ‘D’; the software
allows architects to address the needs of the visually impaired population.
10.1.1 Phases of Testing

There have been four phases of testing a series of software iterations ‘A’, ‘B’, ‘C’
and ‘D’; these tests were used for fine tuning the software for prediction of
visibility for vision impaired people.
2001 to 2003 Phase 1 briefly extensive testing with 35 people with low vision,
in transport hubs, recording the angle, height and distance from 380 objects which
informed Software ‘A’ and identified five factors of interdependent variables. It
was a PC based Windows program, published by Dalke et al. (2010) and predicted
a visibility determining distance line for each VA level. Participants were vision
tested for Visual Acuity, Visual Field and colour vision impairment.
2008 Phase 2 test recruited ten volunteers through eSight, with categories of
vision impairment (V1 - V10) who self-defined their vision using a visual scale
(Grundy et al., 1999; Douglas et al., 2006; Dalke et al., 2010). Participants were
18+ years with a variety of vision impairment conditions such as Retinitis
Pigmentosa, 97% blind with very low field of vision some colour vision,
participants with Nystagmus, night blindness, and others having problems with
bright light, tunnel vision, cone dystrophy, some sight, light sensitive but relying
Visibility Prediction Software: Five Factors of Contrast Perception 95
on rods since birth, peripheral vision only, long sight and reasonable visual acuity
24/60, and another person with wet macular disease. There were no participants from
V1, V7 and V9 (V1 has no possible perception of light).
Phase 2 was conducted in a controlled laboratory testing Software ‘A’ predictions
(Dalke et al., 2010). Each participant was logged as lowest VA group self-defined that
ensured a margin of error for the algorithm. This phase tested distances at which VIPs
were able to see the contrast difference between 2187 greyscale patches, presented in
three different sized patches of ten grey LRVs on ten grey LRV backgrounds of pre-
mixed NCS colours (see Figure 10.1); 150mm2, 300mm2 and 750mm2 patches on a
1800mm by 2400mm background, from up to 10 metres distance away were tested
(Figure 10.2). The LRVs of backgrounds and patches were 5, 10, 21, 27, 40, 53, 62, 71,
82 and 93% LRV positioned randomised at eye level. All greyscales were measured
with a spectrophotometer (xyY) and each test conducted in a daylit room; the distance
on the grid at which the participant observed the patch was recorded.
Figure 10.1. Three different grey patches on a background
Figure 10.2. Diagram of 10m distance testing environment Phase 2
Testing found unacceptable margins of error for small and large objects and the
data was used to assess the accuracy of predictions from Software ‘A’. Software ‘B’
was developed - in the form of spreadsheet 2D ‘lookup’ charts with data lines for all
contrasts that predicted visibility.
2010 Phase 3 was conducted in a lab and explored the boundaries of Software ‘B’.
It extended the data range of the Phase 2, with two participants of V4 and V8 and a new
test distance was added of 20m. Two extra size patches were tested, 1000mm2 and
50mm2 of identical LRVs to Phase 2 which were presented on backgrounds of grey
with an increased test course of 20m. Participants observed the patches placed on three
1500mm2 backgrounds of 5%, 53% and 93% LRV, in randomised sequences. Each test
was in a controlled environment of a lux level range of 200-400 lux (see Figure 10.3).
The distance the patch was observed by the participant on the background was recorded
(see Figure 10.4).
96 Dalke et al.
Figure 10.3. 1000mm2 patch on 1500mm2 background (left). 50mm2 patch on 1500mm2
background (right).
Figure 10.4. Diagram of testing environment and course Phase 3
The results for this phase can be seen (see Figure 10.5a) where a 50mm2 object
of 90 points of contrast did not achieve better than 5m distance perception for a V4
participant, and a V8 perception improved significantly after 20 points of contrast
difference (see Figure 10.5b).This phase of testing informed Software ‘C’, a PC
based DOS program that encompassed the lookup charts developed in Software ‘B’.
Contrast Points Contrast Points

(a) (b)
(a) (b)
Figure 10.5. (a) 50mm and 1000mm Object dimension line and distances seen by V4
participant using results Phase 3. (b) 50mm and 1000mm Object dimension line and
distances seen by V8 participant using results Phase 3.
2010 Phase 4 used a real-world environment for testing which included

variables such as visual noise, and assessed the accuracy of predictions from
Software ‘C’ with six participants V2 to V8. Previous lab tests Phase 2 and Phase 3
had all variables carefully controlled and measured - that is lux levels, dimension,
distance and contrast. Phase 4 test was set in an environment with a mix of visual
noise and stimuli - The Food Store at Kingston University. The six chosen
participants’ visual abilities were logged as self-defined on the visual acuity scale
(Grundy et al., 1999) and were V2, V4, V4, V4, V5 and V8 with a V10 control.
Five locations were selected for testing. Participants were asked to stand at marked
and measured predetermined locations, established by using the Software ‘C’.
Objects at each location were listed on a record sheet with their data, e.g. distance
and lux. The following were noted: smallest dimension of the object, LRV
difference between object and background, lux level. At each location, the
participants were invited to begin by observing the environment in front of them
and move forward and ‘describe their view’. The target object was not singled out
by the researcher; the researcher recorded the distance at which an object in the
environment was clearly perceived by a participant (Figure 10.6).
Software ‘D’ predictions were validated and established the V4+ boundary (V4
to V9 is 93% of VIP) as the software’s minimum default. Final tuning of the charts
from a comparison of Phase 3 and 4 test results, for the algorithm used in Software
‘C’, then ‘D’ was used for an iPhone App, released in November 2010
(www.cromocon.com).
Figure 10.6. Walking in the corridor testing distance perception of objects
10.2 Analysis of Results and Algorithm Development

Previous sections detailed the tests in Phases 2 and 3, obtaining of empirical data,
including Phase 4, used to test the predictions of Software ‘C’ and ‘D’. The
analysis of the results constructed and evaluated a novel algorithm to predict
visibility. A comparison between Phase 2 and 3 results with the predictions from
Software ‘A’ showed significant innacuracies to warrant development of further
software described in the following sections. A new algorithm was developed
using an empirical approach. Results from Phase 2 tests developed Software ‘B’ to
overcome the two problems described below.
98 Dalke et al.
10.2.1 Extending the Range of the Test Data to 20 Metres

Phase 2 test course of 10 metres was increased to 20 metres for Phase 3 with two
participants (V4 and V8). Phase 2 data was extended by ‘typical differences’ where
unconstrained data was analysed to identify the differences between object
dimension ranges, which were then applied to constrained data. The figure below
(10.7) shows results for a V4 subject before (a) and after the adjustments (b).
(a) (b)
Figure 10.7. Test results from Phase 2 and 3, participant A05 (V4), for all contrast points.
Adjusted test data for participant A05 (V4) extended to range of 20m, for all contrast points.
10.2.2 Adjusting the Data to Normalise for a Constant Lux

of 400
A problem with Phase 2 results was optimising the varying lux values between
each participant’s tests (from ~200 to ~1000 lux) for a lux of 400, that would be
effective for a brightly lit task based working environment (Williams, 1999);
results from Phases 2 and 3 were rescaled. Lux-distance relationships under
different conditions of contrast, object dimension, and the VA level were found
using Software ‘A’. Linear approximations to these curves were used to rescale
data points. Figure 10.8 shows this applied to a sample data point (a).
Figure 10.8. Lux-distance lines plotted using data points from Software ‘A’, V4 contrast of
50 and object dimensions of 150, 300, and 750mm. (a) - Phase 2 data point from participant
A05 (V4) at contrast 50, dimension 300mm, and lux of 240 (b) - Data point (a) adjusted for
lux, from 240 to 400 using linear trend-line (distance, contrast 50, and dimension 300mm).
The linear approximation depended on conditions of the data point requiring

adjustment, including the VA level. Figure 10.9 shows results adjusted from 240 to
400 lux can be compared to Figure 10.7b for before/after lux adjustment.
Figure 10.9. Results for participant A05(V4) at 150mm, 300mm, and 750mm dimensions,
adjusted from 240 to 400 lux
In averaging results from Phase 2 and 3 to find a Typical Data Set for Software
‘B’ we focused on the VA 4 level. Results from Phase 2 and 3 were averaged across
all VA levels. A large number of results were used without having to rely on V4 data
sets and allowed us to attenuate the visual anomalies to give a better typical visual
impairment prediction. The visual range was guaranteed by averaging across VA
levels (2+3+6+4+6+3+4+5+8) / 9 = 4.55. Measures of central tendency were
investigated with results from Phase 2, a geometric mean was used; extreme
100 Dalke et al.
anomalous points did not skew these. Central tendency measures were found to have
a geometric mean that gave the best representation for each contrast data set.
10.2.3 Interpolation between Data Points and Final

Adjustments
Linear interpolation is un-representative of a relationship between sight and
distance; techniques were investigated such as logarithmic, exponential, and
polynomial interpolation. Non-Uniform Rational B-splines avoided the oscillatory
nature of previous interpolations as the degree of the curve is fixed independently
of the number of points fitting the curve. The interpolation was also extended in
the contrast direction creating a NURB surface (Piegl et al., 1997) in a 3D visual
space consisting of distance, dimension and contrast axes. The surface describes a
visual threshold - above the surface signifies no perception, and below, signifies
perception. The numbers of data points in the contrast and dimension directions
were used as the basis of knot vectors that gave a control net of 5 x 9 points. The
Cox-de Boor recursion formula (Piegl et al., 1997) was used to define the blending
functions in each parametric direction. Figure 10.10a shows a sample of
interpolated curves superimposed on the same 2D axis.
(a) (b)
Figure 10.10. NURB interpolated curves showing all contrasts for a typical V4 subject (a)
and (b) adjusted NURB interpolated contrast curves for a typical V4 subject
Software ‘B’ was composed of nine lookup charts in Excel of increasing

contrasts from 10 to 90 points. An algorithm was formulated and used these curves
developed in DOS, Software ‘C’. These curves were found to contain anomalies
(seen in Figure 10.10a) where lower contrast lines had higher distances at certain
dimensions; these were resolved by lowering original test data points to fit the
same behaviour as all other data lines, and by readjusting the control points
defining the NURB interpolation between data points. A sample of the revised
curves is shown above (Figure 10.10b). Software ‘C’ was adjusted before porting
the revised algorithm into the app for the iPhone with a graphical user interface,
labelled Software ‘D’.
10.2.4 Testing the Software

Tests in Phase 4 (see Figure 10.6) and results provided data for comparisons of the
predictions of software with participants A02, A06, and A10; A07 and A05 are not
compared, as visibility lines were not established with the participants viewing all
objects at all distances. All the distances in the Phase 4 tests were normalised for
400 lux.
Contrast 60 and 30 comparisons had the most data points at each contrast
stage and represented the majority of the results (See Figure 10.11). The V4
subject (A02) in the charts above includes the percentage deviation from
predicted, with each data point; apart from points labelled (1) which are
considered anomalous. Deviation in both contrasts begins very high with 87%
and 90% but these are deceptive if distance differences between A02 and
prediction are considered. They are similar at all dimensions except for 1000mm
in Figure 10.11a. Considering the differences that can exist in visual impairment
within the same VA level, subject A02 compares well with the predictions for a
typical V4. The two V5 subjects perform lower than the predicted V4, in most of
their data points, when they should be higher. Participant A06 (V5) is especially
low at contrast 30 points.
(a) (b)
Figure 10.11. Comparison of Software ‘D’, typical V4 predictions with Phase 4 test results
at contrast 60 points (a) and comparison of Software ‘D’, typical V4 predictions with Phase
4 test results at contrast 30 points (b)
102 Dalke et al.
10.3 Conclusions
The studies highlighted the inconsistency of participants’ self defined VA level, the
range of particular vision impairments (no two people with identical impairments),
and the unpredictable impact of each individual eye condition on the observation of
targets tested in the ‘real-world’. Not surprisingly participants performed much
worse in busy real-world test locations (Figure 10.6) proving the danger of relying
solely on lab testing for the development of assistive models of visual perception.
However, a secure model has now been synthesised and developed through the
investigation of the broad range of variables for perception. Five factors were
explored and integrated into a practical tool, the app (www.cromocon.com), which
is proving to be a robust and valuable tool for the design of inclusive
environments. Further work is being undertaken with a large cohort of visually
impaired people to extend the empirical work of the study.
10.4 References
ADA (1991) Americans with Disabilities Act. ADA Standards for accessible design.
Accessibility guidelines for buildings and facilities. U.S. Department of Justice
ANSI/IESNA (2007) Lighting and the visual environment for senior living, IES RP-28-07.
New York, US
Barten PGJ (1999) Contrast sensitivity of the human eye and its effects on image quality.
SPIE Press, Bellingham, WA, US
Dalke H, Conduit GJ, Conduit B, Cooper R, Corso A, Wyatt DF (2010) A colour contrast
assessment system: Design for people with visual impairment. In: Langdon PM, Clarkson
PJ, Robinson P (eds.) Designing inclusive interactions. Springer, London, UK
Dalke H (2011) The contrast guide. Cromocon, London, UK
DDA (2004) Disability Discrimination Act. Her Majesty’s Stationery Office, London, UK
Douglas G, Corcoran C, Pavey S (2006) Network 1000: Opinions and circumstances of VIP
people in Great Britain. Visual Impairment Centre for Teaching and Research (VICTAR),
University of Birmingham, Birmingham, UK
EHRC (2010) European Human Rights Commission Equality Act. HMSO London, UK
Goldsmith S (1967) Designing for the disabled. RIBA Publications Ltd, London, UK
Grundy E, Ahlburg D, Ali M, Breeze E, Sloggett A (1999) Disability in Great Britain:
Results from the 1996/97 disability follow-up to the family resources survey. Technical
Report 94. Department of Social Security, Leeds, UK
Harle J, McLannahan H (2008) Visual impairment: A global view. Oxford University Press,
UK
Piegl LA, Tiller W (1997) The NURBS book, 2nd edn. Springer-Verlag, New York, NY, US
Pullin G (2009) Design meets disability. MIT Press, Cambridge, MA, US
Rogers-Ramachandran DC, Ramachandran VS (1998) Psychophysical evidence for boundary
and surface systems in human vision. Vision Research, 38(1): 71-77
Williams W (1999) Footcandles and lux for architectural lighting. An introduction to illuminance,
2nd edn. Available at: http://www.mts.net/~william5/library/illum.htm (Accessed on 28
November 2011)
Part III
Designing Cognitive
Interaction with Emerging
Technologies
Chapter 11
Intrinsic Motivation and Design of ICT for

the Ageing Population
T.S. Goldhaber, P.M. Langdon and P.J. Clarkson
11.1 Introduction
In many countries around the world, the average of the population is rapidly
increasing (OECD, 2006). On average, people require more healthcare as they age
(Blaschke et al., 2009), meaning that in a population in which more people are of
an advanced age, there will need to be a higher level of healthcare provision (Gray,
2005). However, existing medical systems have neither the finances nor capacity to
insure a high standard of care for older individuals (Rechel et al., 2009).
Information and Communication Technology (ICT) systems provide an
increasingly promising platform with which to improve the efficiency and
effectiveness of healthcare, particularly in a preventative context (Blaschke et al.,
2009; Gupta et al., 2009; Rechel et al., 2009). For example, ICT can enable better
communication with carers and medical professionals, allow individuals to better
manage their own care, increase independence for older individuals, and reduce
some of the problems of social isolation associated with old age (Chaffin and
Harlow, 2005; Blaschke et al., 2009). Unfortunately, older people have persistently
been slow to adopt ICT and consequently miss out on many of the benefits that
could drastically improve their well-being (Mynatt et al., 2004; Czaja and Hiltz,
2005; Czaja et al., 2006; Wagner et al., 2010). Poor user interface (UI) design is
partially responsible for some of the problems with ICT adoption in the ageing
population (Hawthorn, 2000, 2007; Dickinson et al., 2005; Czaja et al., 2006).
The problems with interface adoption and use in the ageing population are
partially problems of motivation: older adults are not necessarily motivated to learn
to use existing or new ICT interfaces (OFCOM, 2006). However, while both
motivation in older adults (Lawton et al., 2002; Melenhorst et al., 2006) and the
application of motivation theory to UI design (Williams et al., 2007; Jung et al.,
2010) have been studied to some extent, the direct application of motivation theory
to UI design for the ageing population has not.
Because ICT will continue to present opportunities for improved healthcare and
independence for the ageing population, motivating them to use it is essential. The

106 Goldhaber et al.
purpose of this paper is to begin to investigate how existing theories of intrinsic

motivation might be applied to motivate the adoption and learning of ICT among
older individuals.
11.2 ICT Adoption, Learning, and Use in the

Ageing Population
Because of factors such as declining fluid intelligence, spatial ability, memory, and
attention span, coupled with increased computer anxiety, older users have a much
harder time learning to use new ICT (Hawthorn, 2000). In addition, older people
are less likely to engage in activities that they do not find enjoyable. As explained
by Melenhorst et al. (2001): “Even if they are relatively healthy, older adults
realise their place in the life cycle…they tend to be present-oriented and are
reluctant to spend their time in an unpleasant way” (Melenhorst et al., 2001).
According to Socioemotional Selectivity Theory (SST), imposed time constraints,
including ageing, can lead to a change in overall goals (Carstensen et al., 1999),
and many older users do not see modern ICT as related to their interests or goals
(Hawthorn, 2007). In addition, they may not find the use of an initially difficult
and confusing interface enjoyable and may therefore be reluctant to learn to use it
even if they are aware of the potential benefits (Melenhorst et al., 2006). Learning
problems, anxiety, and lack of familiarity increase the probability that ICT use will
not initially be either enjoyable or useful, leading to reduced adoption and use rates
(Hawthorn, 2000).
Earlier research shows two things:
• There are significant motivational barriers to technology adoption in the
ageing population.
• Motivation theory can be applied to user interface design to increase
motivation in users (e.g. Williams et al., 2007; Jung et al., 2010).
The area that needs to be addressed, therefore, is how motivation theory can be
applied to interface design to increase motivation to adopt and fully learn to use
new ICT. Critically, interfaces aimed at older users must motivate use of the
interface itself, not just the accomplishment of an external goal, so the enjoyment
users get from using the interface must be maintained as well. This makes intrinsic
motivation a crucial area of study.
Intrinsic Motivation and Design of ICT for the Ageing Population 107
11.3 Intrinsic Motivation and Design

The idea of intrinsic motivation is most easily traced back to White (1959), who
summarised a great deal of research supporting the conclusion that not all
motivation can be explained by extrinsic factors. For example, exploratory
behaviour in young animals is rarely extrinsically motivated. Clearly, as White
explained, there is some underlying framework for intrinsic motivation that needs
to be understood. White identified competence, or the ability to affect and cope
with one’s surroundings, as the primary driver of intrinsic motivation (White,
1959). Bandura’s concept of self-efficacy, influenced by White’s work, also relies
heavily on the concept of competence, which is referred to as “performance
accomplishment.” Bandura emphasised that a sense of efficacy early on in an
activity correlates with motivation and, later, persevering through challenges and
setbacks (Bandura, 1977). Intrinsic motivation is critical in design for the ageing
population because they are much less likely to undertake a task if they do not find
it enjoyable (Carstensen et al., 1999; Melenhorst et al., 2001).
To expand on these ideas, the following sections summarise some prominent
theories that deal either in whole or in part with intrinsic motivation.
11.3.1 Self-determination Theory

Self-Determination Theory (SDT) is based on “the hypothesis that there is a set of
universal psychological needs that must be satisfied for effective functioning and
psychological health” (Deci and Ryan, 2008b). Importantly, the fulfilment of these
needs predicts not only overall well-being but also the extent to which motivation
becomes internalised - intrinsic motivation being the mostly highly internalised
form of motivation (Ryan and Deci, 2000b) - and the consequent potential for
learning (Deci et al., 1991). These core needs are defined by Deci and Ryan as
autonomy, relatedness, and competence (Deci and Ryan, 2000).
Autonomy lies at the heart of SDT. Autonomous actions are defined as those
“for which people feel a full sense of choice and endorsement of an activity” (Deci
and Ryan, 2008a). If someone does not feel a sense of volition and control over
their decisions or actions, they cannot be intrinsically motivated because they are
by definition working to fulfil externally set goals or avoid potential imposed
consequences. It is the nature of those goals and consequences that leads to a lack
of autonomy, which in turn can seriously degrade intrinsic motivation.
Relatedness, although not as immediately applicable to interface design, is one
of the key psychological needs specified in SDT. It is defined as “the need to feel
belongingness and connectedness with others,” and is “centrally important for
internalisation” (Ryan and Deci, 2000b).
Competence is “a sense of efficacy” (Przybylski et al., 2010): the feeling
acquired when someone feels that they have the skills necessary to complete a task
or influence their environment. Both rewards and feedback, among other factors
such as an optimal level of challenge, can enhance feelings of competence, but

only if there are sufficient feelings of autonomy - an “internally perceived locus of
causality” (deCharms, 1968; Ryan and Deci, 2000a). The idea of competence as an
important factor in intrinsic motivation was first introduced by White (1959).
11.3.1.1 Cognitive Evaluation Theory

Cognitive Evaluation Theory (CET) is a sub-theory of SDT that focuses
specifically on the causes of intrinsic motivation, particularly the SDT needs of
competence and autonomy (Ryan and Deci, 2000b). Feelings of self-determination
are obviously most affected by the perceived locus of causality and the resulting
feelings of autonomy, but feelings of competence also strongly affect intrinsic
motivation. CET has been tested in over 100 studies and has been found to have
effects in many activities, such as sports and educational environments (Deci et al.,
1999; Ryan et al., 2006). One of the more interesting and consistent experimental
results is the finding that extrinsic rewards actually undermine intrinsic motivation,
meaning that what might promote extrinsic motivation can have the opposite effect
intrinsically (Ryan et al., 1983; Rummel and Feinberg, 1988; Deci et al., 1999;
Ryan and Deci, 2000b).
The concept of competence is complemented and expanded on by the theory of
Flow (Csíkszentmihályi, 1990; Sheldon and Filak, 2008), which is explained in the
next section.
11.3.1.2 Research Implications of SDT and CET

OFCOM (2006) surveyed a large number of older users about their use of
computers and the internet. Those who were interested in learning to use this
technology, but either had not tried or had experienced difficulty, mentioned lack
of skills, benefit perception, and motivation as well as social/environmental factors
as the main barriers to adoption and use. While the fear of becoming socially
isolated due to ICT use points more to the relatedness need of SDT, the problems
with skills clearly point to competence as a key problem. Since ICT use is rarely
mandated or externally rewarded, it is also likely that the need for autonomy may
already be filled for many older users. Therefore, it makes sense that future
research focus first and foremost on filling the competence need.
This approach is supported in the literature. Both White (1959) and Bandura
(1977) cite competence/sense of efficacy as critical to intrinsic motivation,
explaining how acquiring feelings of competence early in an experience can
prevent later negative reactions to errors and promote perseverance in the face of
challenge, both noted issues for older people learning to use ICT. Bandura in
particular emphasises that individuals must have many early successes, which tend
to lessen the negative impact of future failures. UI design in general, therefore,
should focus on creating as many successful experiences as possible, particularly
with initial tasks.
11.3.2 The Theory of Flow

Flow is “optimal experience”: a state of total engagement with or immersion in an
activity (Csíkszentmihályi, 1990). The experience of Flow, by definition, requires
an activity or experience to be enjoyable and is therefore related to intrinsic
motivation (Privette, 1983). Flow in system use has also been shown to lead to
more exploratory behaviour in users (Ghani and Deshpande, 1994). Therefore,
understanding what leads to a Flow experience has the potential to unlock system
properties that could encourage intrinsic motivation in users.
Csíkszentmihályi (1990) identifies three main components of an activity that
lead to Flow. The first is a challenge level that is matched to the skill set of the
user. Second, an activity must have clear and bounded goals such that the person
involved has a direction for their activity. Third, there must be immediate feedback
so that the user or actor knows how they are doing. It is essential in order to
experience Flow that someone knows that they are accomplishing the goals that
have been laid out. (Csíkszentmihályi, 1990).
A separate but nonetheless important feature of the Flow experience is a sense
of control (Fisher, 1978), which although not elaborated in as much detail in
Csíkszentmihályi’s writings is a major component, autonomy, in SDT/CET. Flow
also expands nicely upon the competence and autonomy aspects of SDT, and
competence in particular, (Sheldon and Filak, 2008). This overlap creates both
confidence in the potential of these theories to improve interface design and a
space for experiments and studies testing how these psychological needs and
system properties can be best incorporated into interfaces. In addition,
Csíkszentmihályi reveals that “the more a job inherently resembles a game - with
variety, appropriate and flexible challenges, clear goals, and immediate feedback -
the more enjoyable it will be regardless of the worker’s level of development”
(Csíkszentmihályi, 1990). Indeed, games provide an interesting platform for the
study of intrinsic motivation in addition to indicating that interfaces can provide
flow and intrinsic motivation if they are structured properly.
11.3.3 Other Theories of Intrinsic Motivation and Design

While SDT/CET and Flow are broad theories that deal specifically with intrinsic
motivation, there are other theories that touch on the precursors for intrinsic
motivation and/or design principles for better motivation. This section provides
brief overviews of those theories and their implications for UI design.
11.3.3.1 Theory of Intrinsically Motivating Instruction (TIMI)

In the early 1980s, Thomas Malone studied different educational games, both
interviewing students about their reactions, likes, and dislikes and also modifying
the interfaces in various ways in order to try to isolate motivating properties
(Malone, 1981). As the result of this research, he identified three factors that
appear to facilitate intrinsically motivating instruction within educational
interfaces: challenge, fantasy, and curiosity. Malone’s experimental methods also

provide a good framework for future experimental work in this area.
11.3.3.2 The 2 x 2 Achievement Goal Framework

The 2 x 2 Achievement Goal Framework deals with two dimensions of motivation:
performance/mastery and approach/avoidance. This framework comprises four
achievement goals: “mastery-approach (focused on attaining task-based or
intrapersonal competence), performance-approach (focused on attaining normative
competence), mastery-avoidance (focused on avoiding task-based or intrapersonal
incompetence), and performance-avoidance (focused on avoiding normative
incompetence” (Cury et al., 2006). In general, mastery goals prove more
intrinsically motivating than performance goals, and approach goals more so than
avoidance goals.
11.3.3.3 Positive Design Theory

Ping Zhang (2008a, 2008b) identifies the need to take a motivational approach to
ICT design, explaining that “understand[ing] technology use behaviour and
eventually feed[ing] this understanding into technology design…naturally calls for
a motivational approach because motivation explains what gives behaviour its
energy and direction” (Zhang, 2008b). Zhang’s papers provide good overviews of
how various theories of motivation relate to different design problems.
11.3.3.4 Funology and Gamification

The idea that non-game interfaces should be fun is often overlooked in the field of
Human-Computer Interface (HCI) design (Monk et al., 2002). While it is
obviously not always a desirable feature to have interfaces be fun, particularly
when those interfaces are a means to an end, Draper (1999) identifies three
scenarios in which fun is an important consideration in design: (1) when fun is the
main purpose of the interface, such as a game; (2) when learning is the main
function of the interface; and (3) when “learnability is considered as an important
secondary requirement of software with some other main function” (Draper, 1999).
Monk et al. (2002) proposes the term “Funology” to refer to the study of making
interfaces fun for purposes including but not limited to games, the importance of
which is recognised in the literature, if not pervasively (Rieber et al., 1998; Draper,
1999).
11.3.4 Summary
In addition to Zhang’s design recommendations and the concepts of Funology and
Gamification, four theories relating to intrinsic motivation were presented above:
Self-Determination Theory, Flow, the Theory of Intrinsically Motivating
Interfaces, and the 2 x 2 Achievement Motivation Theory. Interestingly, these
theories all relate to and inform one another to some extent. Figure 11.1 shows the
relationships between the different aspects of each theory.
Figure 11.1. How the intrinsic motivation theories relate to each other
There are many features of these theories that prove incredibly useful for design,
particularly as variables that can be manipulated within experiments. Most
importantly, the idea of competence and challenge is present in all four theories,
although it is expressed and broken down differently in all of them. This is useful
because it is the most relatable to interface design and the varying of challenge
level is one of the most salient features of computer games, which have been
studied extensively. Because of this, because the absence of feelings of
competence has been shown to be one of the primary barriers to the adoption of
ICT, and because feelings of competence have been shown to have effects such as
increasing exploratory behaviour in users (Chaffin and Harlow, 2005) competence
generation must be a primary focus of future UI design research.

Because of lack of prior experience, reduced fluid intelligence, memory capacity
and spatial ability, among other factors, some older individuals can have a harder
time learning new skills than younger users Older users also have low levels of
computer self-efficacy, knowledge, experience, and benefit perception, all of
which reduce the desire to adopt the technology. In addition, older people are much
less likely to engage in activities that they do not find enjoyable. This means that
increasing the intrinsic motivation of users through the application of motivation
theory to design could encourage older individuals to adopt new ICT.
Many theories in social psychology identify elements that lead to intrinsically
motivating experiences. Cognitive Evaluation Theory, a sub-theory of Self-
Determination Theory, identifies competence and autonomy as critical factors for
intrinsic motivation. Flow theory unpacks the experience of competence into the
critical components of appropriate challenge, clear goals, and useful feedback. The
Theory of Intrinsically Motivating Instruction captures how better design can
promote learner engagement. Finally, the 2 x 2 Achievement Goal Framework

shows how different kinds of goals affect intrinsic motivation.
Feelings of lacking competence and self-efficacy have been identified in the
literature as one of the single biggest barriers to adoption, learning and use of ICT
among older individuals. Therefore, design research based on intrinsic motivation
must focus on trying to generate feelings of competence in users and measuring
subsequent changes in intrinsic motivation and learning. Moreover, as competence
is the area of greatest overlap among the theories of intrinsic motivation presented,
this is likely the most important feature on which to focus in future research.
The authors are grateful for the help of Dr Andrew Przybylski in providing and
discussing many valuable sources of information and ideas regarding intrinsic
motivation.
11.6 References
Bandura A (1977) Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84(2): 191-215
Blaschke C, Freddolino P, Mullen E (2009) Ageing and technology: A review of the
research literature. British Journal of Social Work, 39(4): 641-656
Carstensen LL, Isaacowitz DM, Charles ST (1999) Taking time seriously - a theory of
socioemotional selectivity. American Psychologist, 54(3): 165-181
Chaffin AJ, Harlow SD (2005) Cognitive learning applied to older adult learners and
technology. Educational Gerontology, 31(4): 301-329
Csíkszentmihályi M (1990) Flow: the psychology of optimal experience. Harper & Row,
New York, US
Cury F, Elliot AJ, Da Fonseca D, Moller AC (2006) The social-cognitive model of
achievement motivation and the 2x2 achievement goal framework. Journal of
Personality and Social Psychology, 90(4): 666-679
Czaja SJ, Charness N, Fisk AD, Hertzog C, Nair SN, Rogers WA et al. (2006) Factors
predicting the use of technology: Findings from the center for research and education on
aging and technology enhancement (CREATE). Psychology and Aging, 21(2): 333-352
Czaja SJ, Hiltz SR (2005) Digital aids for an aging society. Communications of the ACM,
48(10): 43-44
deCharms R (1968) Personal causation. Academic Press, New York, US
Deci EL, Koestner R, Ryan RM (1999) A meta-analytic review of experiments examining
the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125(6):
627-668
Deci EL, Ryan RM (2000). The “what” and “why” of goal pursuits: Human needs and the
self-determination of behavior. Psychological Inquiry, 11(4): 227-268
Deci EL, Ryan RM (2008a) Facilitating optimal motivation and psychological well-being
across life’s domains. Canadian Psychology-Psychologie Canadienne, 49(1): 14-23
Deci EL, Ryan RM (2008b) Self-determination theory: A macrotheory of human
motivation, development, and health. Canadian Psychology-Psychologie Canadienne,
49(3): 182-185
Deci EL, Vallerand RJ, Pelletier LG, Ryan RM (1991) Motivation and education: The self-
determination perspective. Educational Psychologist, 26(3/4): 325- 346
Dickinson A, Eisma R, Gregor P, Syme A, Milne S (2005) Strategies for teaching older
people to use the World Wide Web. Universal Access in the Information Society, 4(1):
3-15
Draper SW (1999) Analysing fun as a candidate software requirement. Personal and
Ubiquitous Computing, 3(3): 117-122
Fisher CD (1978) Effects of personal control, competence, and extrinsic reward systems on
intrinsic motivation. Organizational Behavior and Human Performance, 21(3): 273-288
Ghani JA, Deshpande SP (1994) Task characteristics and the experience of optimal flow in
human-computer interaction. Journal of Psychology, 128(4): 381-391
Gray A (2005) Population ageing and health care expenditure. Ageing Horizons, 2: 15-20
Gupta M, Chotard L, Ingporsson O, Bastos J, Borges I (2009) Health@home - an e-service
model for disease prevention and healthcare in the home. Electronic Healthcare, 1: 17-
24
Hawthorn D (2000) Possible implications of aging for interface designers. Interacting with
Computers, 12(5): 507-528
Hawthorn D (2007) Interface design and engagement with older people. Behaviour and
Information Technology, 26(4): 333-341
Jung JH, Schneider C, Valacich J (2010) Enhancing the motivational affordance of
information systems: The effects of real-time performance feedback and goal setting in
group collaboration environments. Management Science, 56(4): 724-742
Lawton MP, Moss MS, Winter L, Hoffman C (2002) Motivation in later life: Personal
projects and well-being. Psychology and Aging, 17(4): 539-547
Malone TW (1981) Toward a theory of intrinsically motivating instruction. Cognitive
Science, 5(4): 333-369
Melenhorst AS, Rogers WA, Bouwhuis DG (2006) Older adults’ motivated choice for
technological innovation: Evidence for benefit-driven selectivity. Psychology and
Aging, 21(1): 190-195
Melenhorst AS, Rogers WA, Caylor EC (2001) The use of communication technologies by
older adults: Exploring the benefits from the user’s perspective. Human Factors and
Ergonomics Society Annual Meeting Proceedings, 45(3): 221-225
Monk A, Hassenzahl M, Blythe M, Reed D (2002) Funology: Designing enjoyment
Mynatt ED, Melenhorst AS, Fisk AD, Rogers WA (2004) Aware technologies for aging in
place: Understanding user needs and attitudes. IEEE Pervasive Computing, 3(2): 36-41
OECD (2006) Live longer, work longer: A synthesis report of the Ageing and Employment.
Policies Project. OECD, Paris, France
OFCOM (2006) Older people and communications technology. OFCOM Consumer Panel.
Privette G (1983) Peak experience, peak performance, and flow: A comparative analysis of
positive human experiences. Journal of Personality and Social Psychology, 45(6): 1361-
1368
Przybylski AK, Rigby CS, Ryan RM (2010) A motivational model of video game
engagement. Review of General Psychology, 14(2): 154-166
Rechel B, Doyle Y, Grundy E, McKee M (2009) How can health systems respond to
population ageing? World Health Organization, Copenhagen, Denmark
Rieber LP, Smith L, Noah D (1998) The value of serious play. Educational Technology,
38(6): 29-37
Rummel A, Feinberg R (1988) Cognitive evaluation theory - a meta-analytic review of the
literature. Social Behavior and Personality, 16(2): 147-164
Ryan RM, Deci EL (2000a) Intrinsic and extrinsic motivations: Classic definitions and new
directions. Contemporary Educational Psychology, 25(1): 54-67
Ryan RM, Deci EL (2000b) Self-determination theory and the facilitation of intrinsic
motivation, social development, and well-being. American Psychologist, 55(1): 68-78
Ryan RM, Mims V, Koestner R (1983) Relation of reward contingency and interpersonal
context to intrinsic motivation - a review and test using cognitive evaluation theory.
Journal of Personality and Social Psychology, 45(4): 736-750
Ryan RM, Rigby CS, Przybylski A (2006) The motivational pull of video games: A self-
determination theory approach. Motivation and Emotion, 30(4): 347-363
Sheldon KM, Filak V (2008) Manipulating autonomy, competence, and relatedness support
in a game-learning context: New evidence that all three needs matter. British Journal of
Social Psychology, 47: 267-283
Wagner N, Hassanein K, Head M (2010) Computer use by older adults: A multi-disciplinary
review. Computers in Human Behavior, 26(5): 870-882
White RW (1959) Motivation reconsidered - the concept of competence. Psychological
Review, 66(5): 297-333
Williams GC, Lynch M, Glasgow RE (2007) Computer-assisted intervention improves
patient-centered diabetes care by increasing autonomy support. Health Psychology,
26(6): 728-734
Zhang P (2008a) Motivational affordances: Reasons for ICT design and use.
Communications of the ACM, 51(11): 145-147
Zhang P (2008b) Toward a positive design theory: Principles for designing motivating
information and communication technology. In: Avital M, Boland R, Cooperrider D
(eds.) Designing Information and Organizations with a Positive Lens. Elsevier Science,
Oxford, UK
Chapter 12
A Framework for Studying Cognitive

Impairment to Inform Inclusive Design
E. Jokisuu, P.M. Langdon and P.J. Clarkson
12.1 Introduction
Cognitive impairment is an exceptionally complicated phenomenon, for which no
simple solutions exist. There is no unified definition of what constitutes cognitive
impairment and the terminology is varied. There is no uniform pattern of
impairment occurring with any one medical condition and individual differences
are significant: the same disease can affect people in different ways and the same
condition can cause diverse patterns of impairment in different individuals. Some
medical conditions causing cognitive impairment are chronic, others progressive,
which adds to the difficulty of assessing their impact. In addition, the co-morbidity
of certain medical conditions causes unexpected patterns of impairment. With
certain conditions, such as brain injury, the impact of the impairment depends
entirely on the location and extent of brain damage. (Groome et al., 2006)
Furthermore, the number of people living with cognitive disability is growing due
to population ageing and the improved survival rates for trauma patients (Ghajar,
2000; WHO, 2008). Neuropsychiatric illnesses and traumatic brain injury are
among the leading causes of disability (WHO, 2008).
It is perhaps due to these complicating factors that cognitive impairment has
not been adequately addressed in current research. Related research conducted in
the field of assistive technology has given valuable insight into the design of
devices for specific medical conditions and impairment types (e.g. Gartland, 2004;
LoPresti et al., 2004; Cohene et al., 2007). Design guidelines also exist to outline
some general aspects of product usability, partly aiming to minimise the cognitive
load to users (e.g. Centre for Universal Design, 1997; W3C, 2008).
However, there is no systematic and comprehensive review of design issues
relating to cognitive impairment. Assistive technology products tend to be
developed for a very specific purpose and tailored to each individual, whereas
design guidelines are often too general and fail to provide justification or deeper
understanding of the needs of the people with cognitive impairment. In addition, it
is not easy to estimate the number of people excluded from using a product, as

116 Jokisuu et al.
there are no comprehensive statistical data available about the prevalence and
incidence of cognitive impairment.
Understanding users and knowing their needs and requirements is vitally
important to the success of inclusive design (Carlsson et al., 2002). Still, as
important as user involvement is in the design process, there are particular
challenges when involving users with cognitive impairment. Designers need
support if they are to make design inclusive of such people. One way of supporting
them is to provide information that is accurate, relevant, offers them insight and
inspiration and is presented in a way that makes it easy to apply. This paper reports
the first stage of a study that seeks to develop a model of cognitive impairment
providing such design guidance.
In this first stage, a framework for studying cognitive impairment is developed.
This framework aims to identify those key aspects of cognitive impairment that
designers need to acknowledge whilst also capturing the complexity of the
phenomenon. The purpose of this study is a) to define the cognitive functions
which are relevant to design; b) to categorise the medical conditions which can
cause impairment in these functions; and c) to develop a framework which can be
used to systematically collect information about cognitive impairment as it applies
to design. This is the first stage of a larger study which aims to model cognitive
impairment in a way that translates medical and psychological information into
design guidance.
This paper is organised as follows: Firstly, the research method is described
briefly. Then the development of categories for both cognitive impairment and the
medical conditions causing it is presented, followed by a description of the
framework created based on these categories. Next an example of using the
framework is given. The paper concludes with a discussion of the work so far and
future steps of the research.
12.2 Research Method

Cognitive impairment can be diagnosed through the medical condition causing it.
In order to identify and define the different types of cognitive impairment and the
medical conditions behind them, interviews with experts on neuropsychology and
occupational therapy were conducted. The purpose of these interviews was to
create two categorisations: one of cognitive impairment and another of medical
conditions causing cognitive impairment. In addition, the purpose was to discuss
ways to assess the links between the medical conditions and the types of
impairment.
12.2.1 Interview Technique

Semi-structured interview was chosen as the method for collecting information
from the subject matter experts. As the area of interest was not very well defined,
instead of pre-defined specific questions, a list of general themes for discussion
A Framework for Studying Cognitive Impairment to Inform Inclusive Design 117
was covered during the interview to explore the issues of interest. The interview
outline is presented in Table 12.1. Each interview lasted for about one hour. The
interviews were recorded using a voice recorder and then transcribed. A consent
form was signed by all interviewees.
Table 12.1. Interview themes
What do you think of these categories of medical conditions? What do you think of
these categories of cognitive functions?
How would you assess the strength of the link between the medical condition and the
cognitive functions? You can think of the strength in terms of how often do these co-
occur or how typical it is for a person who has this disorder to have this cognitive
problem.
What is the most representative/most characteristic disease/disorder of each disease
category? What are the most debilitating functions/conditions? Other diseases typical
of the category?
Is it possible to categorise disorders based on impaired cognitive functions?
Is it possible to describe the strength of the link between a disease and a function?
Is it possible to link brain areas to specific functions?
12.2.2 Participants
Altogether four people were interviewed: two neuropsychiatrists with both clinical
and academic experience and two occupational therapists specialising in cognitive
impairment. The interviewees were chosen based on their relevant expertise in both
academic and clinical settings. Each of them had over 10 years of experience of
working with people with cognitive impairment.
Participant 1 was a specialist in developmental neuroscience. His area of
expertise covered different forms of learning disabilities and included Alzheimer’s
disease, autism and depressive disorders as well as birth defects and congenital,
developmental intellectual disabilities. He also had an interest in the citizenship of
people with learning disabilities and their role in the society.
Participant 2 was a neuropsychiatrist specialising in epilepsy and autism. His
research interests included neuropsychiatric features and symptoms of neurological
diseases as well as the biological processes and brain areas associated with certain
neuropsychiatric diseases.
Participants 3 and 4 were occupational therapists. They were interviewed
simultaneously. One of them was an expert in traumatic brain injury, the other in
learning disability.
118 Jokisuu et al.
12.3 Categorisation of Cognitive Impairment

The International Classification of Functioning, Disability and Health (ICF) by the
World Health Organisation was used as the basis for the identification and
grouping of cognitive functions. There are many ways of assessing the functional
capability of an individual, with differing definitions of functions and capability,
which is why an international well-recognised classification - ICF - was chosen. It
is a classification of human functioning and disability, and it covers body functions
and structures and their impairments as well as the potential activity limitations and
participation restrictions following from them. (WHO, 2001)
The classification of cognitive functions presented in ICF did not require many
changes to be applicable to this study. Based on the interviews, the categorisation
of cognitive functions seems relatively standard, although ‘the challenge is to find
categories which cover everything and are recognised by people’ (Participant 3).
However, the occupational therapists emphasised that there is no standard way of
categorising diseases and functions and there are many different ways of labelling
them. In addition, people from different professions interpret the terms in different
ways. For example, occupational therapists and doctors understand the terms very
differently. The categorisation based on ICF and the categorisation based on the
interviews are presented in Table 12.2.
12.4 Categorisation of Medical Conditions

In order to get a comprehensive view of the medical conditions with potential
impact on cognitive capability, existing and established classifications were used.
Two of the most widely used classifications are the International Classification of
Diseases (ICD-10) by the World Health Organization and the Diagnostic and
Statistical Manual of Mental Disorders (DSM-IV) by the American Psychiatric
Association (Mezzich, 2002). These were used together as the basis for developing
the categorisation of medical conditions appropriate for this study.
While both classifications have been developed as diagnostic aids, there are
significant differences between them. ICD-10 has a basic grouping of epidemic
diseases, constitutional diseases, local diseases arranged by site, developmental
diseases and injuries, whereas DSM-IV focuses on diagnosing mental health issues
with some correspondence to the mental disorders of ICD-10 (WHO, 2007; APA,
1994).
The categorisation of medical conditions changed considerably as a result of
the interviews. As Participant 1 mentioned, the medical classifications are ‘not the
most logical way’ of categorising diseases and disorders:
‘Medical conditions can be categorised in a number of different ways, and the way
of classifying them does not matter as long as there is some logic to it.’
His suggestion was to have broader categories and present examples of diseases in
each category. Table 12.3. describes the changes from the categorisation in ICD-
10/DSM-IV to the adapted categorisation based on the experts’ opinion.
Table 12.2. Developing a categorisation of cognitive functions based on ICF (WHO, 2001)
Categorisation of Cognitive Categorisation Adapted for the

Impairment in ICF Framework
Perception Visual perception Perception Visual perception

Auditory perception Auditory perception
Tactile perception Visuospatial perception
Watching Olfactory, gustatory and
Listening tactile perception
Attention Sustaining attention Attention Focusing attention
Shifting attention Shifting attention
Dividing attention Dividing attention
Sharing attention Sharing attention
Memory Sensory memory Memory Short-term memory
Short-term memory Long-term memory
Implicit memory Retrieval of memory
Episodic memory
Semantic memory
Retrieval of memory
Thinking Abstraction Thinking Pace, form, content and
Organization and control of thinking
planning Abstraction
Time management Organization and planning
Cognitive flexibility Time management
Insight Cognitive flexibility
Judgment Insight
Problem solving Judgement
Making decisions Problem-solving
Decision-making
Language Comprehending Language Reception of language
messages Expression of language
Producing messages Integrative language
Conversation functions
Calculation Calculation
Learning Copying Learning Copying
Rehearsing Rehearsing
Learning to read, write Learning to read, write
Acquiring skills Acquiring skills
Psycho- Psychomotor control Psycho- Psychomotor control
motor Quality of psychomotor motor Quality of psychomotor
functions functions functions functions
Sequencing complex Sequencing complex
movements movements
Vestibular functions
Proprioceptive function
120 Jokisuu et al.
Table 12.3 Developing a categorisation of medical conditions based on ICD-10

(WHO, 2007) and DSM-IV (APA, 1994)
Categorisation of Medical Conditions Categorisation Adapted for the

in ICD-10/DSM-IV Framework
• Infectious diseases, e.g. Tuberculosis Disorders of development affecting
• Neoplasms cognition
• Diabetes • Down’s syndrome
• Autistic spectrum disorders, e.g.
• Dementia autism, Asperger’s syndrome
• Delirium • Cerebral palsy
• Disorders due to psychoactive • Specific learning difficulties, e.g.
substance use dyslexia
• Schizotypal disorders
Neurodegenerative disorders
• Mood disorders
• Alzheimer’s disease
• Neurotic and stress-related disorders
• Parkinson’s disease
• Disorders of adult personality and • Multiple Sclerosis
behaviour
• Mental retardation Acquired brain injuries
• Disorders of psychological • Neoplasms
development • Traumatic brain injury
• Infections, e.g. Meningitis,
• Behavioural and emotional disorders
Encephalitis
• Inflammatory diseases of the central
nervous system, e.g. Encephalitis Vascular disorders
• Systemic atrophies primarily affecting • Stroke
the central nervous system, e.g. • Cardiovascular accidents
Huntington’s disease
• Extrapyramidal and movement Episodic disorders affecting the
nervous system
disorders, e.g. Parkinson’s disease
• Epilepsy
• Degenerative disorders of the nervous • Delirium
system, e.g. Alzheimer’s disease • Migraine
• Demyelinating diseases of the central
nervous system, e.g. Multiple sclerosis Illnesses of other systems affecting the
• Episodic and paroxysmal disorders, brain
e.g. Epilepsy • Diabetes and other metabolic and
endocrine disorders
• Polyneuropathies and other disorders
of the peripheral nervous system Psychiatric disorders affecting
• Cerebral palsy and other paralytic cognition
syndromes • Schizophrenia
• Cerebrovascular diseases • Depression
• Congenital malformations of the • Bipolar disorder
nervous system
• Head injuries
12.5 Categories as a Matrix

Cognitive impairment can be diagnosed through the medical condition causing it.
Therefore, building the link between the medical conditions and the cognitive
functions they impair is essential. For this purpose, the categorisations described
previously were used to create a matrix with medical conditions on rows and
cognitive functions in columns (see example in Figure 12.1). The idea was to use
the matrix to reveal patterns of impairment characteristic of certain medical
conditions by mapping the links between medical conditions and cognitive
functions. The mapping was done by assessing the strength of the connections
between any given medical condition and the cognitive functions it typically
impairs. The strength assessment had a scale of 1-3, with one denoting a weak link
and three a strong link. Zero was used to indicate that there is no known link
between the medical condition and the cognitive function. This matrix forms the
basis of the framework.
In this stage of the study, the matrix was used as a tool in the interviews to get a
better understanding of the potential usefulness and applicability of the
categorisations, to elicit opinions and experiences and to guide the experts in
thinking in the field of cognitive impairment as a whole. In addition to getting
valuable feedback on the categories, interviews also offered a chance to start
learning about the possible patterns linking medical conditions and cognitive
impairment.
All the participants thought that the matrix was an interesting tool and a new
way of approaching the area. In theory there should be a pattern of signs and
symptoms evident that identifies each disease; this pattern is used to diagnose the
disease. With accurate enough data, patterns could be seen but the data would
probably be ‘noisy’ (Participant 2).
Perception
Language
Attention
Learning
Thinking
Memory
Psycho-
motor
Down’s syndrome
Dyslexia
Autism
Brain injury
Schizophrenia
Figure 12.1. Example of using the framework (from the study data). The darker the shade,
the more significant the impairment.
All of the participants highlighted potential challenges of the approach. The

main problem would be that the data collected with the matrix would not
122 Jokisuu et al.
necessarily be meaningful due to individual variations and the huge range of

severity in conditions: the severity of the impairment depends on the progression of
the disease and varies with the stage of the illness. Another challenge is those
medical conditions, for example stroke and brain injury, which could cause
problems in all of the cognitive functions. As Participant 2 emphasised:
‘It must be remembered that the matrix would only provide an average, a good
approximation perhaps but very variable. […] The biggest challenge is probably
going to be the immense amount of information that needs to be extracted,
synthesised and summarised.’
12.6 Using the Framework

After the initial series of interviews, a brief validation study was conducted to
ensure the categorisation was as accurate as possible. Of the four interviewees,
three reviewed the adapted categorisations to make sure they correctly reflected the
participants’ views. To get a better idea of how the framework would work in
practice, Participant 2 filled in the matrix based on his professional experience;
meaning, he assessed the significance of each medical condition in terms of
causing cognitive impairment (Figure 12.1). His experience was that it takes a
considerable amount of time and effort to complete the matrix. The main
advantage of the framework as an information collection tool lies, in his opinion, in
its flexibility: it is possible to look at cognitive impairment as a whole or drill
down to details of specific medical conditions.
‘It would help designers to understand where the individual’s strengths are by
understanding the weaknesses and where tailoring of a product would be
advantageous. The key is to find a useful way of conveying the information and not
get impeded by the complexities.’
After this initial study, the framework has been used to collect information
regarding various types of cognitive impairment and the medical conditions which
cause them. This work is still ongoing but the framework seems to work quite well
in that it allows a multifaceted view into the phenomenon of cognitive impairment:
a general overview as well as more detailed information, as needed. The
framework can also be used to identify the most important areas of cognitive
impairment; that is, the most common and severe types of impairment which
designers should focus on. An example of the framework is presented in Figure
12.1 in which the darker shades indicate areas where impairment is particularly
significant. As a result, these would be the areas that designers should pay attention
to. The example is discussed in more detail in Jokisuu et al. (2011). At this stage,
however, the data is incomplete, as analysis is still ongoing. Incorporating the
expert opinion of more participants into the framework will improve the reliability
and validity of the results. Similarly, differences in participants’ assessments will
highlight areas requiring further examination.
12.7 Conclusions
The framework presented in this paper can be used to systematically analyse the
various types of cognitive impairment and the medical conditions that cause them.
It is based on medical and disability classifications which were modified according
to experts’ opinions. Through interviewing the experts, it became evident that there
is a need for an information tool which would translate the sometimes complicated
medical material into a more accessible format that would be relevant to designers.
However, there are several potential problems with the framework which were
highlighted during the study. The individual differences and varying levels of
impairment, particularly in progressive diseases, are extremely difficult to capture
with the framework. Terminology is not clear and the definitions of functions and
disorders are not understood the same way, even when a well-established
international classification is used. It should also be noted that impairment in a
function does not necessarily result in an inability to perform an action; people
develop compensatory strategies to cope in everyday life. These are not captured
by a somewhat simplistic theoretical model. In addition, at this first stage of the
study, only four people participated in the development of the framework, which
might limit the validity of the work; however, this issue will be addressed in later
stages when the framework is further developed and applied in a larger-scale study.
The strength of the framework is that it offers designers a multifaceted view of
the complexity of cognitive impairment: an overview of the general issues as well
as the details and background. In the next stages of the study, the framework will
be used to collect information that will provide a more complete picture of
cognitive impairment, including examples of assistive technology and existing
design guidance, stories of people living with cognitive impairment as well as a
general idea of the number of people affected by various types of impairment. The
framework will then be further developed with designers to see which elements are
most useful and applicable to them.
12.8 References
APA (1994) Diagnostic and statistical manual of mental disorders, 4th edn. American
Psychiatric Association, Washington DC, WA, US
Carlsson G, Iwarsson S, Stahl A (2002) The personal component of accessibility at group
level: Exploring the complexity of functional capacity. Scandinavian Journal of
Occupational Therapy, 9(3): 100-108
Centre for Universal Design (1997) Universal Design Principles. Available at:
http://www.design.ncsu.edu/cud/about_ud/udprinciples.htm (Accessed 21 October 2011)
Cohene T, Baecker R, Marziali E, Mindy S (2007) Memories of a life: A design case study
for Alzheimer’s Disease. In: Lazar J (ed.) Universal usability: Designing computer
interfaces for diverse users. John Wiley & Sons Ltd, Chichester, UK
Gartland D (2004) Considerations in the selection and use of technology with people who
have cognitive deficits following acquired brain injury. Neuropsychological
Rehabilitation, 14(1/2): 61-75
Ghajar J (2000) Traumatic brain injury. The Lancet, 356(9233): 923-929
124 Jokisuu et al.
Groome D, Brace N, Dewart H, Edgar G, Edgar H, Esgate A et al. (2006) An introduction to

cognitive psychology. Processes and disorders, 2nd edn. Psychology Press, Hove, UK
Jokisuu E, Langdon PM, Clarkson PJ (2011) Modelling cognitive impairment to improve
universal access. In: Stephanidis C (ed.) 6th International Conference on Universal Access
in Human-Computer Interaction (UAHCI 2011), Orlando, FL, US
LoPresti E, Mihailidis A, Kirsch N (2004) Assistive technology for cognitive rehabilitation:
State of the art. Neuropsychological Rehabilitation, 14(1/2): 5-39
Mezzich JE (2002) International surveys on the use of ICD-10 and related diagnostic
systems. Psychopathology, 35(2-3): 72-75
W3C (2008) Web Content Accessibility Guidelines WCAG 2.0. Available at:
http://www.w3.org/TR/WCAG20/ (Accessed 21 October 2011)
WHO (2001) International classification of functioning, disability and health. World Health
Organization, Geneva, Switzerland
WHO (2007) International statistical classification of diseases and related health problems,
10th edn. Available at: http://www.who.int/classifications/apps/icd/icd10 (Accessed 21
October 2011)
WHO (2008) The global burden of disease: 2004 update. Available at:
http://www.who.int/entity/healthinfo/global_burden_disease/GBD_report_2004update_ful
l.pdf (Accessed 21 October 2011)
Chapter 13
Interactive Error Correction Using

Statistical Language Models in a Client-
server Interface for Editing Mathematical
Text
D. Attanayake, G. Hunter, E. Pfluegel and
J. Denholm-Price
13.1 Introduction
Learning and using mathematical notation poses particular difficulties for people
with various disabilities, partly due to its wide range of symbols and rather
complicated layout. These pose great challenges, often affecting the educational
and career opportunities of people who are visually impaired or have limited (or
no) use of their hands or arms. Assistive systems to alleviate these difficulties
would be of considerable benefit to such groups of people.
In this paper, we provide an overview of the development of the new client-
server architecture of TalkMaths, a web-based speech interface system for dictating
and editing mathematical text in electronic documents. First, we describe the
current web-based architecture of TalkMaths and its new multi-modal features.
We then discuss the process of building a data corpus by “crawling” through
source codes of carefully selected web sites containing mathematical equations at
the appropriate level. Each equation in this data set is then converted to the
corresponding sequence of mathematical ‘keywords’ in order to use the data to
build language models based on unigram, bigram and trigram statistics (Young,
1996). Finally we explain how these language models will be used within the new
version of TalkMaths to give it automatic predictive or corrective abilities to
improve its performance, usability and user satisfaction levels.

126 Attanayake et al.
13.2 Related Systems

So and Watt (2006) used the predictive power of statistical language models in
their system for recognising hand-written mathematical symbols using optical
character recognition. They extracted mathematical expressions from a public e-
Print server, arXiv.org, which covers a broad range of mathematically based
subjects at research level, and analysed statistics on the most commonly used
expressions. The key difference between our work and that of So and Watt is that
even though the domain (mathematics) is essentially the same in both, the input
methods and related problems are quite different as we are dealing with spoken
input as opposed to hand-written characters. Also, the level of mathematics in
question differs significantly between the two pieces of work, as we focus on
relatively elementary levels of “high school” mathematics whereas So and Watt
consider research level mathematics . There are various other existing systems that
enable the dictation of mathematics such as MathTalk (Stevens et al., 1997), Math
Speak & Write (Guy et al., 2004) and CamMath (Elliott and Bilmes, 2007). The
first two use as their underlying system commercially available mathematical
editors, which are not freely available. Also they offer somewhat limited methods
for navigating and editing each mathematical expression. The Math Speak & Write
system is freely available, although it only allows a rather small set of
mathematical vocabulary and seems to require use of somewhat non-intuitive
editing paradigms. Bernareggi and Brigatti (2008) developed a system for speech
input of mathematics targeted at blind users, but it is currently not widely available
and only supports spoken input in Italian. Hanakovič and Nagy (2006) also
propose a speech input mathematical editor for the visually impaired. However,
their system requires use of XHTML voice (X+V) technology which is restricted to
a limited range of web browsers, such as the Opera web browser.
13.3 Current TalkMaths Architecture

The TalkMaths system is primarily aimed at helping people with various
disabilities (Wigmore et al., 2010a), particularly those with limited or no use of
their hands or arms. The current TalkMaths prototype system uses an “Application
Proxy architecture” (Attanayake et al., 2011), based on a client-server model, and
is a web-based editor for mathematical text. Our client application is a Mozilla
Firefox browser and requires a speech recognition engine to be installed on the
client machine. A web server and a Google apps service are used to realise web
logic (controlling the behaviour of the application independent from any speech
input) and speech logic (parts of the logic that deal with interpreting input in the
form of speech commands) respectively. In contrast to the earlier desktop version
of TalkMaths, the new version now benefits from several advantages offered by
web-standards (well-established technologies for creation and interpretation of
web-based content). First, and perhaps most importantly, it offers easy access to a
large number of users and, secondly, computationally expensive processes like
parsing a sequence of commands can be carried out on a fast web server
Interactive Error Correction Using Statistical Language Models 127
(Gruenstein et al., 2008). This web-based architecture also allows TalkMaths to

store an archive of a user’s interactions with the system, especially the speech
commands used. This can be used for a variety of purposes. Users can retrieve
work they have previously done using the system, without being tied to a particular
machine. The archived data can also be used for analytical purposes, to build a
speech data corpus that can be used both to improve the general centralised
statistical language model (see Section 13.4 below) and/or to develop specialised
language models specific to particular topics or individual users. TalkMaths is also
now a multi-modal system that can accept input by speech and/or keyboard and
mouse.
13.4 Building Statistical Language Models from

Web Data Sources
Our approach to building a substantial mathematical database has been evolving
continuously over the last few years. Our first attempt at obtaining data on the
ways in which mathematical equations are spoken was based on part of the British
National Corpus (BNC) which consisted of transcriptions of conversations from
school and college mathematics classes (Wigmore et al., 2009; Wigmore, 2011).
The vocabulary (number of different words) found in this dataset was 4,355 (the
total number of words was 123,821) and the perplexity of the statistical model
obtained was rather high, due to both the relatively large vocabulary and the high
proportion of non-mathematical words and “chat” in the conversations, making the
data somewhat unpredictable. Our second attempt at modelling mathematical text
was to manually populate a dataset using trigonometric equations from some
mathematical text books (Wigmore et al., 2010b; Wigmore, 2011). Not
surprisingly, both the vocabulary (102 distinct words) and the dataset overall (the
total number of words was 7,857) were smaller than for the BNC-based data. The
statistical language models built from the second (trigonometry) dataset had much
lower perplexity scores, indicating that this data was more predictable.
Interestingly, consistent with earlier studies (Hunter and Huckvale, 2006), the
results of these first two attempts at creating statistical language models from
mathematical material confirmed that, for data of a given type, if the training
dataset is increased in size, the perplexity of the resulting language model
decreases and hence its predictive power is increased. With this in mind, in the
current study we have attempted to create a much larger high quality dataset of
mathematical expressions on which to base new statistical language models. We
identified and “crawled” a handful of public “tutorial” web sites containing
mathematical equations at a similar level of complexity to that which TalkMaths is
currently capable of processing.
For this work, we developed a “web-crawler” that can identify LaTeX or
MathML content within the source code of the web site. This mathematical content
is then extracted into a database. A “filtering script” was then applied to remove
display instructions from LaTeX code and illegal characters from the equations.
Finally, a LaTeX/MathML to “spoken mathematical expressions” converter was
designed and developed using the yapps2 (Patel, 2009) parser generator. The
converter introduced additional keywords “begin” and “end” for denoting sub-
sections (such as fractions or square roots) within the “linearised spoken
descriptions” of each mathematical expression, to create a word string which
would be identical to the “correct” way in which a TalkMaths user would dictate
that expression. We then used the Carnegie Mellon Statistical Language Modeling
(CMU SLM) Toolkit (Clarkson and Rosenfeld, 1997) to build statistical language
models based on an initial sample of our corpus (3,194 mathematical expressions
containing a total of 61,479 words with a vocabulary of 100 words). The process of
building these models was analogous to that used in our earlier studies, which have
been described elsewhere (Wigmore et al., 2010b; Wigmore, 2011).
Table 13.1. Cross-validation perplexity calculations on statistical language models of 3,194

spoken mathematical expressions using CMU Language Modelling Toolkit
Vocabulary
Training set Test Set Training words Test words Perplexity
(words)
subsets [1-9] subset10 54907 6572 7.07 100
subsets[1-8 & 10] subset9 54968 6511 7.17 100
subsets[1-7 & 9-10] subset8 55294 6185 7.11 100
subsets[1-6 & 8-10] subset7 55688 5791 7.31 100
subsets[1-5 & 7-10] subset6 55172 6307 7.25 100
subsets[1-4 & 6-10] subset5 55597 5882 7.74 100
subsets[1-3 & 5-10] subset4 55340 6139 7.01 100
subsets[1-2 & 4-10] subset3 55805 5674 7.65 98
subsets[1 & 3-10] subset2 55177 6302 7.53 100
subsets[2-10] subset1 55363 6116 7.02 100
Initial experiments to investigate the quality of this data confirmed the trends noted
by previous studies (Hunter and Huckvale, 2006; Wigmore et al., 2010b;
Wigmore, 2011). The results from this new study are summarised in Table 13.1.
Interestingly, the predictive power of the models based on the sample tested was
better than those of both our earlier studies (Wigmore et al., 2010b; Wigmore,
2011). There are several possible reasons for this improvement. The vocabulary is
still relatively small and the training datasets used in our latest study are
considerably larger than the trigonometric dataset used in Wigmore et al. (2010b).
Perhaps the most important reasons could be the higher quality and increased
amount of training data.
We will describe below how we will use this statistical model in the context of
our new web-based TalkMaths architecture.
13.5 Speech Error Correction

In our usability studies (Wigmore et al., 2010a), we noted that commercial speech
recognition engines are not primarily designed for the dictation of mathematics.
The speech recogniser’s error rates are normally higher for dictating mathematics
in contrast to “everyday” natural language. However, together with extensive user
training and restricting the vocabulary, this problem can be substantially solved.
While some researchers try to correct recognition errors at the parser level using
error recovery strategies, others rely solely on the language models associated with
the speech recognition engine to solve the problem. Since the users should
themselves know exactly what they meant to dictate in the first place, we believe a
collaborative effort between the user and the system, in the form of an interaction
that is natural to the user, is required to carry out a repair on an erroneous input.
Studies in Suhm, et al. (1996) also discuss interactive error recovery strategies. In
addition, They deduce “user preference factors” in interactive error recovery such
as the “naturalness of” the interaction, interpretation accuracy and the time it takes
for a user to provide an input and for the system to interpret it. We will address
these issues in our error correction environment to minimise the time and effort a
user has to spend to complete a correction with ease.
13.6 Design of Interactive Prediction Mechanism

Since speech recognition is carried out at the client side, the statistical language
model can be used to predict what will be said next, preferably also at the client
side. To realise this, the statistical language model is transmitted to the client as a
one-off transaction when the client first connects to the web server. This model is
then used by our companion application, residing in the client machine, which
captures and passes recognition results from the speech recognition engine on to
the browser. If the statistical language model produces some alternative predictions
such as a ranked list of most likely alternatives to or continuations of the
recognised word sequence, these will be sent to the remote server for parsing, and
the results sent back to the client in the form of MathML expressions, which will
be displayed as “alternative” mathematical equations in two-dimensional format
on the screen. Note that the original utterance from the speech recognition engine
will always be transmitted to the server for parsing and consequently displayed to
the user as the “primary” mathematical expression on the screen. The user can
subsequently check the appearance of the primary expression and, if this is not
what he originally intended, he can correct it, either by editing the primary
expression or by selecting the appropriate choice from the set of displayed
alternatives, by speech or using keyboard and mouse. The process of our
“interactive predictive alternative” system is illustrated in Figure 13.1 overleaf.
SLM at server
Send SLM to client
Display maths Wait for next Correction or

(including alternatives if utterance or navigation
present) instruction
No
Maths?
Yes
Parse main utterance No

or instruction Alternative(s)?
Yes
Parse alternative(s)
Figure 13.1. Flowchart of TalkMaths “predictive alternative” system
13.7 Conclusions and Future Work

We believe our “predictive alternative” system (described above) driven by a
statistical language model, together with natural editing strategies, can significantly
improve user satisfaction levels for TalkMaths in the future. Furthermore, by
reducing speech recognition and parser errors in TalkMaths, disabled people such
as repetitive strain injury (RSI) sufferers should find TalkMaths to be an intuitive
and useful tool. We will carry out an empirical evaluation of our multi-modal
system with both disabled and non-disabled users in order to answer several key
scientific questions relating to its design and usability. We are also investigating
more sophisticated parsing methodologies and editing paradigms with a view to
making further improvements to TalkMaths.
13.8 References
Attanayake D, Hunter G, Pfluegel E, Denholm-Price J (2011) Architectures for speech-
based web applications. In: Proceedings of the International Conference on Semantic E-
Business and Enterprise Computing (SEEC 2011), Kingston upon Thames, UK
Bernareggi C, Brigatti V (2008) Writing mathematics by speech: A case study for visually
impaired. In: Proceedings of the 11th International Conference on Computers Helping
People with Special Needs (ICCHP 2008), Linz, Austria
Clarkson P, Rosenfeld R (1997) Statistical language modeling using the CMU-Cambridge
Toolkit. In: Proceedings of the 5th European Conference on Speech Communication and
Technology (Eurospeech 1997)
Elliott C, Bilmes JA (2007) Computer based mathematics using continuous speech
recognition. In: Proceedings of the CHI 2007 Workshop on Striking a C[h]ord: Vocal
Interaction in Assistive Technologies, Games, and More, San Jose, CA, US
Guy C, Jurka M, Stanek S, Fateman R (2004) Math speak & write, a Computer program to
read and hear mathematical input. Electrical Engineering and Computer Sciences
Department Technical Report, University of California, Berkeley, CA, US
Gruenstein A, McGraw I, Badr J (2008) The wami toolkit for developing, deploying, and
evaluating web-accessible multimodal interfaces. In: Proceedings of the 10th International
Conference on Multimodal Interfaces (ICMI 2008), Chania, Greece
Hanakovič T, Nagy M (2006) Speech recognition helps visually impaired people writing
mathematical formulas. In: Proceedings of the 10th International Conference on
Computers Helping People with Special Needs (ICCHP 2006), Linz, Austria
Hunter G, Huckvale M (2006) Is it appropriate to model dialogue in the same way as text? A
comparative study using the British National Corpus. In: Proceedings of the 2006
European Modelling Symposium, London, UK
Patel A (2009) Parsing with yapps. Available at: http://theory.stanford.edu/~amitp/yapps/
(Accessed 25 August 2011)
So CM, Watt SM (2006) Determining empirical properties of mathematical expression use.
In: Proceedings of the 4th International Conference on Mathematical Knowledge
Management (MKM 2005), Bremen, Germany
Stevens RD, Edwards, ADN, Harling, PA (1997) Access to mathematics for visually
disabled students through multimodal interaction. Human-Computer Interaction 12(1):
47-92
Suhm B, Myers B, Waibel A (1996) Designing interactive error recovery methods for
speech interfaces. In: Proceedings of the CHI 96 Workshop on Designing the User
Interface for Speech Recognition applications, SIGCHI, Vancouver, Canada
Wigmore A (2011) Speech-based creation and editing of mathematical content, PhD,
Wigmore AM, Hunter G, Pfluegel E, Denholm-Price J, Binelli V (2009) “Let them
TalkMaths!” Developing an intelligent system to assist disabled people to learn and use
mathematics on computers through a speech interface: the TalkMaths and VoiceCalc
systems. In: Proceedings of the 5th IEEE/AAAI International Conference on Intelligent

Environments, Barcelona, Spain
Wigmore AM, Hunter G, Pfluegel E, Denholm-Price J, Colbert M (2010a) “TalkMaths
Better!” Evaluating and improving an intelligent interface for creating and editing
mathematical text. In: Proceedings of the 6th International Conference on Intelligent
Environments, Kuala Lumpur, Malaysia
Wigmore AM, Pfluegel E, Hunter G, Denholm-Price J, Colbert M, Attanayake D (2010b)
Evaluating and improving the TalkMaths speech interface for dictating and editing
mathematical text. In: Proceedings of 5th European Workshop on Mathematical and
Scientific E-Contents, Salamanca, Spain
Wigmore AM (2011) Speech-based creation and editing of mathematical content. PhD
Thesis, Kingston University, London, UK
Young S (1996) Large vocabulary speech recognition: A review. IEEE Signal Processing
Magazine, 13(5): 45-57
Chapter 14
Understandable by Design: How Can

Products be Designed to Align with User
Experience?
A. Mieczakowski, P.M. Langdon, R.H. Bracewell,
J.J. Patmore and P.J. Clarkson
14.1 Introduction
Understanding users increases the likelihood that the final designed product will
meet the needs of heterogeneous people (Kouprie and Sleeswijk Visser, 2009).
However, the process of learning to understand users and their experiences requires
qualitative research (Mattelmäki and Battarbee, 2002) and a structured investment
of time that can be achieved through the development of a manageable design plan
in the early stages of design (Yang and Epstein, 2005). Although the fields of
accessibility and usability have made significant advances in the last decade in
facilitating the design of products and services that satisfy the needs of different
users and are easy to use, the majority of design companies still fail to
acknowledge users’ needs early on in the design process (Gulliksen et al., 2003).
This is because of, among other things, tight schedules and a limited budget (Dong,
2005).
Studies by Ricability (2001) show that products designed for people with
reduced capability are also easier for everyone else to use. Many products,
however, are designed as if every user were an agile 25-30 year old professional
(Benktzon, 1993) and, therefore, are largely inaccessible and unusable for less
capable users (Keates and Clarkson, 2003).
User-centred design approaches such as Inclusive Design have been developed
in order to help designers expand the boundaries of product usage for as many
people as possible by iteratively attuning product design to the needs of
heterogeneous users from the beginning of the design process (Keates and
Clarkson, 2003). There are moral and financial benefits associated with application
of the Inclusive Design philosophy, as products which are designed in line with
users’ capabilities are believed to improve customer satisfaction (Hewer and

134 Mieczakowski et al.
James, 1998), which in turn allows companies which value good design to achieve
high growth (Keates and Clarkson, 2003).
The work presented in this paper discusses the ways in which design companies
can effectively adopt the inclusive design ethos and addresses the development of a
modelling approach for supporting designers in determining early stages in the
design process, whether specific product features evoke similar understanding and
responses among the users and the designers of those features.
14.2 Inclusive Product Design

To design more inclusively, designers must “understand the users, their cognitive
behaviour, attitudes and the characteristics of their work tasks” (Gulliksen et al.,
2003). Likewise, Preece et al. (2002) argue that a product should be based on
users’ needs, requirements and the actions that they will perform when performing
tasks and that it should be tested iteratively.
Chamberlain et al. (2011) cite the case of a large global communications
company which serves customers in 170 countries, BT plc, to illustrate how user-
centred design can be effectively adopted in the development of products and
services. The study by Chamberlain and colleagues is based on five
recommendations: (1) it is vital for senior management to support the principles of
user-centred design and identify a champion or champions who will promote this
change within an organisation; (2) this design approach will be adopted more
successfully if it is positioned as a design that improves the lives of everyone and
advocates that small changes can make a big difference; (3) rethinking the design
and accommodating it to the needs of heterogeneous users makes for good social
responsibility and lucrative commercial decisions; (4) the adoption of inclusive
design requires guidelines and tools for supporting the act of designing, as well as
providing training on how to use different support materials; and (5) the most
effective way of designing intuitive interactions in products and services is by
having direct contact with a range of different users.
The subsequent sections of this paper focus on the fourth element of the above
recommendations for adopting the principles of inclusive design, i.e. they describe
the development of a modelling approach for helping designers align their
understanding and use of products with that of users.
14.3 Human Understanding and Experience with

Designed Products
Human understanding is often perceived as a model and the theoretical construct of
a mental model has, for over sixty years, been studied at length by researchers from
a wide range of disciplines (Craik, 1943; Johnson-Laird, 1983; Norman, 2002;
Payne, 2008). Mental models are believed to be heavily influenced by previous
knowledge and experience with similar products (Langdon et al., 2007).
Understandable by Design 135
One of the biggest advocates of mental models, Norman (2002) posits that
accessible and usable products can be designed by means of matching the
designer’s conceptual model of a product with the users’ mental models of that
product through the use of the product interface. However, so far there is a
significant lack of a simple-to-use support method which would guide designers in
representing and comparing their understanding and intended use of a given
product with the users’ understanding and actual use of that product. The aim of
such a support method would ideally be to help designers “reduce the number of
actions and decisions that users have to make in order to reach their goals”
(Langdon and Thimbleby, 2010).
14.4 Modelling Approach for Aligning Design

Intent with User Experience
Since designers are in general comfortable and familiar with interpreting abstract
visual representations such as models and diagrams because they are often
involved in their daily work (Andreasen, 1994), the new inclusive design support
tool proposed in this paper is a data model containing information gained from user
observations and engineering and design insights. The key requirements for the
proposed modelling approach are: (1) to be used early in the design process to
identify potential problems and increase the chances of producing a more inclusive
product; (2) to assist designers in gathering information about users’ understanding
and behaviour in relation to products and comparing it with their own design
intent; (3) to be easy to understand, implement and use; and (4) to be a semantic
modelling representation capturing key pieces of information about users’
understanding and use of a given product.
This paper proposes the Goal-Action-Belief-Object (GABO) modelling
approach for designers to enable the assessment and comparison of designers’ and
users’ understanding and usage of everyday products. This approach works on the
premise that the contents of people’s minds (i.e. their knowledge, theories and
beliefs) should be explored by designers in order to better understand users’
behaviour in relation to products (Payne, 2008). However, it should be noted that
this approach does not focus on the structure of the human mind and does not
represent information about the actively changing parts of mental models. Instead,
it represents the current and static contents of users’ mental models during
interaction with a given product.
The GABO approach encompasses both elicitation and representation
(Klimoski and Mohammed, 1994). Fundamentally, the GABO approach aims to
encourage designers to pay greater attention to users’ understanding and use of
products through representing: (1) goals that users want to achieve during
interaction with products; (2) actions that they exert on product interfaces; (3)
beliefs about product features that they bring from previous interactions with other
products to interactions with new products; and (4) understanding of the impact of
their actions on a product’s functional objects.
During the use of the GABO approach, designers are recommended to create
three types of model: (1) an engineering model of a given product indicating how
its different parts interact with one another; (2) a designer model of a given product
(in order to be compared with the engineering model to see what features should be
mounted on the top of the underlying functional parts); and (3) a number of
individual user models encompassing how users understand and use different
product features (in order to be compared with the designer model). The designer
model and the user models are of the highest importance as they are to be
compared for compatibility, while the engineering model acts as a reference to the
designer model. During the trial of the GABO approach, the designer, engineering
and user models constructed using the approach have been tested on an example of
a coffee maker with an elaborate interface shown in Figure 14.1.
Figure 14.1. A coffee maker with an elaborate interface
This coffee maker has a dual functionality: its left side makes an espresso type
of coffee and its right side is for making a café type of coffee. The espresso side
has such main features as: a start/stop button, a 1 and 2 cup espresso selection
button, a button for steaming hot water and milk, a blue LED display (shared with
the café side), a scroll button for setting time, an ‘OK’ button (shared with the café
side). The café side has the following main parts: a start/stop button, a button for
selecting the programmable mode of the coffee maker, a toggle for selecting strong
or normal café strength, a blue LED display (shared with the espresso side), a
scroll button for setting time and choosing the automatic mode of the coffee maker
(shared with the espresso side) and an ‘OK’ button.
Thirty users, two designers and two engineers were consulted during the trial
with this coffee maker. Information from the users was obtained by observing them
using the coffee maker. Information from designers and engineers was gathered
from discussions based around a semi-structured interview technique. The
engineers and designers consulted worked for the same design company, were
highly educated and had over two years of experience in product design. The thirty
users were of different ages, capabilities and levels of education. Based on the
information they provided, the engineering, designer and user models were created
by the use of the Design Rationale Editor (DRed) software (Aurisicchio and
Bracewell, 2009). The semantic coding language used to create the designer and
user models was composed of different verb and noun combinations depending on
which type of element—goal (verb + noun), action (verb+ing + noun), belief
(noun + to+verb), object (noun)—it was describing. This is in line with the work
of Andreasen (1994) who argues that “the modelling needs to be made semantic to
function like a language”. Subsequently, the constructed GABO models were
compared for similarities and differences to the designer model and each individual
user model using an appropriate algorithm (discussed as follows).
The GABO approach stipulates that any two models (a designer model and an
individual user model) can be compared based on: (1) presence of the same nodes
in the two models and (2) connectivity between two given nodes in the two
models. The comparison procedure can be carried out using an algorithm from set
theory that measures similarity between graphs with common node and edge sets
(Goldsmith and Davenport, 1990). This algorithm is used for measuring both the
presence of nodes and the connectivity between nodes in the designer model and
the individual user models, with the designer model acting as the standard model
against which each user model is checked for compatibility. The assumption is that
by using the GABO approach’s algorithm, designers will be able to make close
estimates of the compatibility of their intended goals, actions, beliefs and objects
regarding product usage with the goals, actions, beliefs and objects of
heterogeneous users.
14.4.1 The GABO Algorithms for Checking Presence and

Connectivity of Nodes
The algorithms for checking the presence and connectivity of nodes in the designer
model and each individual user model are both inspired by the correlation
coefficient, which assesses the shared number of nodes/edges between two sets of
nodes/edges standardised by a measure of the total number of nodes /edges from
both sets. The presence algorithm and the connectivity algorithm consist of the
following steps: (1) examine the designer model (D) and enumerate all
nodes/edges sequentially; (2) examine an individual user model ( U n ) for each
node/edge and identify, by checking the type of each node and keyword, the type
and keywords of the two nodes it connects to, whether it has previously been
enumerated in the designer model (D): if not, then allocate a new number to it
following the numerical sequence from the designer model (D); (3) count the
nodes/edges shared between the designer and user sets of nodes/edges; (4) count
the number of nodes/edges in the designer model; (5) divide the number of shared
nodes/edges between the two sets of nodes/edges by the total number of
nodes/edges from the designer model; and (6) add any nodes/edges that are unique
to an individual user model.
The presence algorithm is: P ( D, U ) = (| N D ∩ N U | / | N D |) + UN U

where P = value between 0 and 1, where
• 0 = total incompatibility between nodes in the designer and user models
• 1 = 100% compatibility between nodes in the designer and user models
• Any value between 0 and 1 = the degree of compatibility between nodes in
the designer and user models (e.g. if P ( D,U ) = 28/88, then the
compatibility level = 0.31 or 31%)
D = designer model
U = user model
N D = the set of nodes in the designer model
N U = the set of nodes in the user model
N D ∩ N U = the set of all nodes that are members of both N D and N U
UN U = the set of unique nodes in the user model.
The presence algorithm assumes that two nodes are the same (one from the
designer model and one from the user model) if they belong to the same element
type, for instance the belief element, and contain the same semantics. The
implementation of the presence algorithm in DRed uses the transclusion function
(Kolbitsch and Maurer, 2006). It should be noted that no distinction is made in the
weightings of certain types of nodes and edges as the approach only checks for the
presence or lack of similar concepts in the understanding of designers and users
and does not estimate the significance of different aspects of their understanding.
In other words, each node or edge present is allocated the same number. In a
similar fashion, each missing node or edge has the same number deducted.
The connectivity algorithm is: C ( D, U ) = (| E D ∩ EU | / | E D |) + UEU
where C = value between 0 and 1, where
• 0 = total incompatibility between edges in the designer and user models
• 1 = 100% compatibility between edges in the designer and user models
• Any value between 0 and 1 = the degree of compatibility between edges in
the designer and user models (e.g. if C ( D,U ) = 30/114, then the
compatibility level = 0.26 or 26%)
D = designer model
U = user model
E D = the set of edges in the designer model
EU = the set of edges in the user model
E D ∩ EU = the set of all edges that are members of both E D and EU
UEU = the set of unique edges in the user model.
The connectivity algorithm assumes that two edges are equal if they join two
nodes that are exactly the same element-wise and semantically in the designer
model and the user model.
The assumption with regard to the comparison procedure of the GABO approach
is that it requires both computer assistance and human input. Computer assistance is
needed for comparing the count of the GABO approach’s elements (nodes and
edges) and generating a value between 0 and 1, which indicates the degree of
presence and connectivity compatibility of each individual user model with the
designer model. Human input is required for creating data structures for each model,
comparing the engineering model with the designer model, as these two types of
models are structured in a different way, and verifying at the end of the automatic
comparison whether the elements in the designer model and the user models have
been compared correctly by the software system.
Once the GABO comparison algorithm was developed, it was programmed into
DRed software using the TCL scripting language by DRed’s proprietor, Rob
Bracewell, based on the research requirements. This code enabled a comparison of
compatibility between the designer model and thirty individual user models of any
product in the DRed software (discussed as follows).
14.4.2 Usage Trial of the GABO Modelling Approach

During the trial of a coffee maker with an elaborate interface, distinct differences were
found between designers’ and users’ understanding and usage of the function for
activating coffee making. In particular, to activate the coffee maker users first had to set
a clock and all the participants failed to envisage that a clock would be the first feature to
be used on the coffee maker interface. Because of this cumbersome design, two
participants were not able to activate the coffee making function, and the rest spent a
significant amount of time (an hour was the longest) trying to figure it out by pressing all
buttons on the interface including buttons on the espresso side. Another difficulty that
most of the people had was with finding the position of the café water chamber. Since
the coffee maker is of dual functionality, designers thought that a logical solution would
be to design a water chamber for each of the functional sides. However, most users
assumed that there was only one water chamber. Four users could not locate the water
chamber on the café side when making filter coffee and poured water in the water
chamber on the espresso side. Moreover, one participant poured water over the filter
holding the coffee grounds, which resulted in all the water dripping out.
Figure 14.2 shows a comparison screen produced at the end of the comparison
procedure of the designer and individual models of thirty users of a complex-to-use
coffee maker in the DRed software. This figure shows scores of compatibility for
presence of nodes in the designer and thirty user models and scores of compatibility for
connectivity between nodes in the designer and thirty user models. In addition, it shows
a count of nodes and edges that are unique to individual user models. The instances of
unique nodes are: new goal (NG), new action (NA), new belief (NB), new object (NO),
repeated node (REP), rejected node (REJ) and misunderstood function (MF). The
instances of unique edges are: new edge (NE) and repeated edge (RE).
Overall, Figure 14.2 indicates that the average index of compatibility for the
presence of nodes is approximately 86% and most users had a substantial number of
unique nodes as they had difficulty with figuring out that a clock function had to be set
before the coffee making function could be activated.
NODES Unique Nodes EDGES Unique Edges
Figure 14.2. GABO evaluation of a coffee maker with an elaborate interface
In addition, since this coffee maker was of dual functionality, when most users were
confused about the functionality of the café side they moved over and used the
functionality of the espresso side thinking that it would help them activate the function
on the café side. Hence an increased number of new actions, new beliefs, repeated nodes
and rejected nodes. The overall score for the presence of nodes is about 85% and there
are large numbers of new edges and repeated edges for linking the increased number of
nodes. For example, the model of user 7 scored 68% on the compatibility scale for the
presence of nodes and it contained eight new actions, three new beliefs, twenty-five
repeated nodes and ten rejected nodes. It needs noting that this user was not able to
activate the coffee making function and gave up after having pressed all interface
buttons with no expected feedback from the coffee machine. Although this user was not
able to make coffee using a complex coffee maker, the overall count of compatibility for
this user is relatively high as this person, while employing a ‘trial and error’ method,
managed to use a number of relevant interface features. A high count for new actions,
new beliefs, repeated nodes and rejected nodes is due to this user pressing a large
number of unrelated buttons to activate the coffee making function. Regarding the score
of compatibility for the presence of edges, this user scored 61% and her/his model
contained forty new edges and twelve repeated edges for connecting eight new actions,
three new beliefs, twenty-five repeated nodes and ten rejected nodes.
14.4.3 Evaluation of the GABO Modelling Approach

The usefulness and effectiveness of the GABO approach was evaluated with eight
designers from a range of small and large organisations based in the UK during two
five hour workshop sessions.
Designers, aged between 29 and 52, were asked to work on two redesign tasks, one
of which required them to redesign a complex-to-use interface of a household product
(either a toaster or a coffee maker) using a method of choice and the other task required
them to redesign an interface of one of the aforementioned two products using the
GABO approach. When the tasks were completed, each designer was asked
individually to fill out an evaluation questionnaire composed of a number of
quantitative and qualitative questions. Overall, the designers marked, on average, point
5 on a 7-point scale indicating how useful the GABO approach was in identifying and
capturing users’ understanding and the problems users encounter during product use.
This procedure was mirrored when investigating designers’ opinion regarding their
understanding of product functionality compared with the understanding of users, the
result being point 5.5 on the scale. Likewise, indicating ease-of-use, the designers on
average gave the GABO approach a score of 4.3. In addition, five designers believed
that the GABO approach helped them to produce a better design than their alternative
method of choice, while three designers said that they would need more time to use the
GABO approach to determine as to whether it was better or worse than the alternative
method. Furthermore, it should be noted that there was general fear among most
evaluators that the GABO approach would require a lot of time and effort in training
and actual usage to prove useful and they would not always be able to use it to its full
potential as most projects are time and cost restricted.
14.5 Conclusions
This paper discussed the ways in which companies can effectively adopt the principles
of inclusive design and the development of a modelling approach for helping designers
align their understanding and use of products with those of users.
Since there seems to be a gap in the availability of an easy-to-use and pragmatic
technique for representing and comparing designers’ and users’ understanding and
usage of everyday products, this paper proposes the GABO approach that aims to
bridge that gap for designers. The GABO approach consists of four stages in which
designers need to: (1) analyse the engineering model of a product to better understand
how different product parts interact with one another; (2) create a designer model of
that product using appropriately annotated GABO elements and compare it with the
engineering model to see what features should be mounted on the top of the underlying
functional parts; (3) investigate how different individuals understand and use product
features, create several individual user models using the GABO elements annotated in
the same semantic style as their counterpart elements in the designer model; and (4)
compare the designer model with individual user models using a simple algorithm from
set theory, check the degree of compatibility between the designer model and the user
models and make appropriate design decisions relating to the inclusivity of future
product features.
The results of the evaluation study with eight designers show that the use of the
GABO modelling approach is feasible in real design situations as designers found it
beneficial in representing their own and users’ understanding and use of products and
subsequently comparing matches and mismatches in the understanding of the two
groups. The research on the GABO approach continues and the next stage is to evaluate
its usefulness and effectiveness in creating novel designs.
14.6 References
Andreasen MM (1994) Modelling: the language of the designer. Journal of Engineering Design,
5(2): 103-115
Aurisicchio M, Bracewell RH (2009) Engineering design by integrated diagrams. In: Proceedings
of the International Conference on Engineering Design, Stanford, CA, US
Benktzon M (1993) Designing for our future selves: The Swedish experience. Applied
Ergonomics, 24(1): 19-27
Chamberlain M, Esquivel, J, Miller, F, Patmore, J (2011) BT’s adoption of customer centric
design. Applied Ergonomics Special Issue, in press
Craik KJW (1943) The nature of explanation. Cambridge University Press, Cambridge, UK
Dong H (2005) Barriers to inclusive design in the UK. PhD Thesis, Cambridge Engineering
Design Centre, University of Cambridge, Cambridge, UK
Goldsmith TE, Davenport DM (1990) Assessing structural similarity of graphs. In: Schvaneveldt
RW (ed.) Pathfinder associative networks, Ablex Publishing Corporation, Norwood, NJ, US
Gulliksen J, Goransson B, Boivie I, Blomkvist S, Persson J, Cajander A (2003) Key principles for
user-centered systems design. Behaviour and Information Technology, 22(6): 397-409
Hewer S, James L (1998) Realising potential: Two complementary views from the RSA, London.
In: Placencia-Porrero I, Ballabio E (eds.) Improving the quality of life for the European Citizen:
Technology for inclusive design and equality. IOS Press, London, UK
Johnson-Laird PN (1983) Mental models. Harvard University Press, Cambridge, MA, US
Keates S, Clarkson PJ (2003) Countering design exclusion: An introduction to inclusive design.
Springer, London, UK
Klimoski R, Mohammed S (1994) Team mental model: Construct or metaphor? Journal of
Management, 20(2): 403-437
Kolbitsch J, Maurer H (2006) Transclusions in an HTML-based environment. Journal of
Computing and Information Technology, 14(2): 161-173
Kouprie M, Sleeswijk Visser F (2009) A framework for empathy in design: Stepping into and out
of the user’s life. Journal of Engineering Design, 20(5): 437-448
Langdon PM, Lewis T, Clarkson PJ (2007) The effects of prior experience on the use of consumer
products. Universal Access in the Information Society, 9: 209-225
Langdon P, Thimbleby H (2010) Inclusion and interaction: Designing interaction for inclusive
populations. Interacting with Computers, 22(6): 439-448
Mattelmäki T, Battarbee K (2002) Empathy probes. In: Proceedings of the Participatory Design
Conference, Palo Alto, CA, US
Norman DA (2002) The design of everyday things. Basic Books, London, UK
Payne SJ (2008) Mental models in human-computer interaction. In: Jacko JA, Sears A (eds.) The
human-computer interaction handbook: Fundamentals, evolving technologies and emerging
applications, Taylor & Francis, NY, US
Preece J, Rogers Y, Sharp H (2002) Interaction design: Beyond human-computer interaction. John
Wiley & Sons, NY, US
Ricability (2001) Inclusive design: Products that are easy for everyone to use. Disability Rights
Commission, London, UK
Yang MC, Epstein, DJ (2005) Study of prototypes, design activity and design outcome. Design
Studies, 26(6): 649-669
Part IV
Design for Inclusion

Chapter 15
Design Advisor: How to Supply Designers

with Knowledge about Inclusion?
E. Zitkus, P.M. Langdon and P.J. Clarkson
15.1 Introduction
In an ideal scenario accessibility issues such as legibility, usability and associated
cognitive load, with respect to users of different age groups and impairments,
would be considered whenever a new product was created. This means that
designers would be challenged by the relation between users’ capability loss and
the features of the new product. In this context, understanding what would cause
exclusion from use of the products could support designers to make decisions
resulting in more accessible products. Consequently, different approaches to
supply designers with knowledge about inclusion have been studied throughout the
last decades. This chapter analyses how interactive design advisors could provide
knowledge about inclusion according to the design activity.
15.1.1 The Need for Integrating User Requirements

A way of dealing with accessibility problems is by supplying designers with
techniques to evaluate the physical interactions that will occur between users and
new concept designs. Currently, there is a range of techniques to evaluate
accessibility in new product development, some of which are listed below:
• user-centred techniques: well known techniques that enable designers to
understand user’s needs and develop empathy to them through users’
participation in the design process (Norman, 2002). Examples of co-
designing methods are seen in the works of Sanders (2000) and Rode et al.
(2004). The user-centred method is also applied in user trials, user
observations and more recently, user theatre (Newell et al., 2006);
• Designer trial techniques: these techniques enable designers to experience
physical restrictions that they are unfamiliar with. Although there are
different types of simulation apparatus such as Third-Age Suit, Age

146 Zitkus et al.
Explorer and Simulation Toolkit, all of them follow the same principle of
restricting the motion and sensorial capabilities (Hitchcock et al., 2001;
Cardoso and Clarkson, 2007).
• Virtual techniques are computer-based techniques in which assessments
occur through software applications. They can be integrated to CAD
models, through task performance simulation like HADRIAN,
INCLUSIVE CAD, and in the future VERITAS and VICON. There are
also tools that simulate vision and hearing capability loss by loading
sounds or images, like the Impairment Simulator or tools that calculate the
exclusion caused by capability demand, like the Exclusion Calculator
(Marshall et al., 2004; Macdonald et al., 2006; VERITAS D4.1.3_v2,
2010; Clarkson et al., 2007).
The advantages and disadvantages of adopting these techniques in the industrial
context were discussed in a previous study and a summary of each approach is
shown on Table 15.1 (Zitkus et al., 2011)
Table 15.1. Different techniques and their integration into design in the industrial context
Process integration Interface Results
User trials / User Observation of real users Inspiring. Exclusion is not quantifiable.
observation and/or getting their Re-assessing the product is an issue
feedback after the trial. due to the sample selection.
Early in the conceptual Inspiration is limited as the design
phase, through similar Observation of
Self observation teams do not represent a wide range
themselves.
products, or later through of people.
Third-Age Suit / rapid prototypes. Designers observe
Inspiring. Exclusion is not quantifiable.
Age Explorer themselves with physical
Re-assessing the product means to
restrictions or different
Simulation Toolkit wear the suit again.
levels of restrictions.
Simulation of functional
Exclusion is limitedly quantifiable due
demand on lower limb
INCLUSIVE CAD the range of tasks and the focus on
muscles, hip and knee
physical capabilities.
During the conceptual joints.
phase through CAD Exclusion is limitedly quantifiable due
HADRIAN Virtual interaction with
models. the range of tasks & users’ database.
user avatars. Exclusion could be quantifiable due to
VERITAS project
a broad anthropometric and
VICON project Virtual simulation. capabilities database.
During the conceptual Simulation of some of
Impairment Exclusion is limitedly quantifiable due
phase through new vision and hearing
Simulator the focus on sensorial capabilities
concept images. capability loss
Early in the concept Virtual interaction with a
Exclusion Dependent on the knowledge of the
phase, through task range of applicable
Calculator designer.
analyses. tasks.
Although the range of tools cited above seems to be a good support for
assessing the accessibility of industrial products, they are barely used in the
industry as part of the design process. This indicates that either the techniques to
assess the product accessibility do not work in tandem with the design process or
that there may be inadequacies on the application interface. Consequently,
techniques that integrate the user’s needs to the design activity and its process in
the industrial context are still needed.
Design Advisor: How to Supply Designers with Knowledge about Inclusion? 147
15.1.2 Proposing an Interactive Technique to the Design

This chapter analyses and discusses the study carried out with experienced
industrial designers (average of 10 years of experience) in two well recognised
design consultancies in the United Kingdom. The design consultancies are
specialised in product design, research and innovation within a broad range of
industrial sectors and clients. The study was based on observations of five
designers at work and recorded interviews with six other designers. At the end of
each interview there was a demonstration of an interactive technique that informs
the designers about inclusion. The study is part of a research project that aims to
consider the design activity and its process while developing new interactive
methods.
The principle of the interactive technique proposed is to help designers to
obtain information about the accessibility afforded by the new concept design
during the time the designers are creating it. The demonstration enabled the
designers to talk about interactive ways to evaluate accessibility that are informed
by the design activity. The designers were asked to comment on the technique.
The interactive technique was built into Google SketchUp software through
Ruby programme language, both open free-sources available on internet. The
demonstration emulated the design of a simple medicine pack and proposed an
interactive way to check the legibility of the letters on the pack. The demonstration
(illustrated in Figure 15.1) followed the sequence below:
• designing the box (with colour and material);
• adding the text (with font size and style);
• setting the ambient light;
• setting the reading distance;
• selecting visibility test in the inclusive design tool in the tools window;
• receiving advice from a window that pops up on the screen, which
describes the range of population excluded from reading the text in the
pack and advice regarding font size, style and background/foreground
colour contrast.
The findings are presented and discussed in the following sessions.
15.2 Interactive Interface for Inclusive Design

The idea of having an interactive adviser built into a computer graphic tool had a
positive response. The proposed interactive interface built into a computer graphic
system (in this case Google SketchUp) was associated with tools to evaluate the
mechanical or structural aspects of three dimensional (3D) models. For instance,
mould flow analysis, stress-strain and FEM analysis were mentioned.
Many of the comments, however, highlighted that the use of this kind of tool
would be guided mainly by the needs of the client, instead of the designer’s work
routine.
148 Zitkus et al.
Figure 15.1. The interactive interface demonstrated to the designers in the study
“So, if you can turn around and say “we’ve got stuff built into this programme
could you check it (accessibility) later and tell if the design is suitable” then they
(the client) would probably love it…. It could be useful if you haven’t got access to
users and if the company need this kind of data… as we know, not every company
can get access to, then it is probably better if they’ve got some kind of evaluation
….”
Some designers mentioned other options that they use to evaluate the
accessibility of new concept designs. Among them are information brought to the
project by the clients themselves, guidelines, user opinions and self- evaluations.
15.2.1 Briefing, Research and Guidelines

All the interviewees mentioned a similar process that happens at the initial stage of
the design activity. Usually, designers are guided by a brief that is a source of
information of the new product’s functionality, the components and their
respective size, manufacturing, environmental considerations and by information
from the potential user.
It was highlighted that user data in the brief is limited to market information,
and thus designers have to manage their time and budget to get user information
from other sources. The designers mentioned that often some research starts taking
place at this stage. This is described in detail in another study (Zitkus et al., 2012).
The accessibility, however, “is not something that always got designated time
within the process...”. Nevertheless, according to the designers, if it is part of the
project requirement to consider accessibility, then they would normally look for
ergonomic data from books, tables and internet, or they would look for
specifications in guidelines.
In fact, the responses indicate that designers mainly rely on guidelines, though
their comments also highlighted that they find the information on these sources
deficient and sometimes incompatible with their needs. They mentioned that they
balance the deficiency of the guidelines by including some live assessments, such
as self-evaluations and user-trials:
“a lot of it I would say is based on common sense, we tend to tell to ourselves
what is legible or not… I think lots of it comes with experience, the way our minds
work, it becomes obvious if something is small and illegible… There are standards
which drive how large a piece of text should be, you can print things out in various
sizes and get feedback from the user group.”
“I think lots of time that happen, that stuff (accessibility considerations) comes
from experience… you are making subconscious decisions of what is good and bad
accessibility. So, I think most of that coming from trying and testing ways of doing
things.”
“My approach would be to print out or to create different variants of the design
and then just test that with people, just quickly get people there, or taking it to
people, just going to Tesco, just finding people being demographically, just find
some people of specific age group or if you got the profile of the user you will
need.”
Although the possibility of incorporating users in accessibility tests was
mentioned, all interviews stressed that user observation or user trials only take
place if the research allowance covers that or if the client defines this as a method
of assessment. The user’s involvement in the process therefore occurs very rarely.
“…even when I worked in companies that project things specifically for the
elderly, it was rare in the extreme anybody who was elderly would be involved in
the process… the users were not part of the process…”
It is important to underline that the designers highlighted that the user’s needs,
such as those related to accessibility and usability are only one part of the
requirements that they have to deal with. They emphasised that design is a
compromise activity, where decisions are made all the time and costs are involved
in every option taken.
“… the product is not only the users themselves, we have to consider who
gonna assemble it by making assembly easier, how it is built, how it, if it gonna
need maintenance, let's say, filters have to be changed, something like that, we
have to consider how can we make that appropriate, easier or not, and at these
days, go through the product going out and ending his life and needing to be
recycled. So, to assemble, the materials, all that side of things…”
“… to be honest, in many stages there are simple costs and practicality costs
but all the primary driver before we get things like accessibility.”
150 Zitkus et al.
The designers also had some opinions about the interface of the interactive
technique demonstrated. Their opinions bring to the light suitable places for a
possible application and also the effectiveness of the information supplied to them.
15.2.2 Cognisant Applications for the Design Practice

According to some interviewees an interactive adviser for graphic designers
regarding the legibility of text, icons or any other graphic interface should be built
for graphic designers into their respective tools:
“I think most of the designers agencies use Illustrator or Photoshop so if it is
incorporated on that it would be part of the tool you are used to using rather
having to learn something… you don’t have to spend ages learning a new
programme;”
For other designers the idea of having all the analysis in 3D systems would
facilitate their work. The interface could be two dimensional, but it should be
integrated to a CAD software (such as SolidWorks and Pro-Engineer).
“So, if there was a tool which can plug into CAD which allow us to set up
different ergonomic parameters, certain aspects of dexterity, or visual acuity, or
things like that, and then to be able to see and to get some kind of feedback with
that, so that would be useful, I can’t see why it wouldn’t be… If the principle is to
design in CAD, I think you should have as much information as possible in that
environment.”
“Two dimensional could be useful in 3D software not in terms of graphic stuff,
but I guess a 2D ergonomic database, like knee height. That easily gives you that
information, it would be quite useful.”
15.2.3 The Relevance of the Inclusion Information

Another aspect of the inclusive interactive interface that caught the designers’
attention was the percentile of the population excluded. For them there is always a
target market that guides the design activity. Advices based on the percentile of the
entire population are generally not significant in such cases where a portion of the
population is not part of the target market. The exclusion value that appeared on
the pop-up window is therefore ineffective.
“In nearly all of the products we work on it, the portion of the population
might be not applicable to that, because they might be irrelevant, the usability. We
would assume that some people will be excluded when you see that product; and
that is acceptable for that product be successful, still it is from client perspective…
I’m not sure, when you get a result like 7.3% of the population will be excluded the
question is “which part of the population is it talking about?” Because if it is
excluding 80% of people over 75 years old females from North East, so that is
really important if the product is aiming that; those people in the North East in
England. The detail of that is useful, that is what we need to know I guess.”
The next section discusses the study and its relation with previous work in the
design activity, as well as the relationship of the design activity with the
accessibility evaluation techniques.
15.3 Deficiencies of User Information

The interactive technique proposed and demonstrated to the designers supported
the study in understanding the way accessibility evaluation techniques are selected
and the role of the client in considering the user’s requirements. The designers
highlighted ways in which the interface could be applied in computer graphic
systems as well as the relevance of the information provided. In the following
sections the outcomes are compared with past research and discussed in detail.
15.3.1 Sources of User information

Past studies had already mentioned that time and budget constrain the adoption of
user trials in the industrial context (Gill, 2009; Goodman-Deane et al., 2010). The
findings described in the previous section confirm that direct involvement with
users rarely happens in commercial projects.
Users are consumers and their profiles are brought to designers in the form of
target markets, generally consisting of a demographic view of potential consumers.
User requirements, such as those related to the diverse range of capabilities among
the potential consumers, are not necessarily part of the market information
provided by the client.
Any extra information regarding user data should come from the designer’s
own research. This research, however, is limited due to the project budget and the
client’s procedures, as a result of which user requirements are generally restricted
to ergonomic tables available in books or on the internet. However, these tables
generally do not consider a wide range of people, including the elderly and the
disabled.
In situations where accessibility is part of the design requirements, then
designers make use of guidelines to comply with the users’ needs. In this case, the
findings of the study confirm the deficiency implicit in the guidelines that were
already mentioned in past studies.
Cardoso and others (2003) noted that the new possibilities and the range of
features of a new concept design are not covered by guidelines. Past studies also
indicated that the broader the scope of the guideline the less it supports the design
activity (Burns et al., 1997; Law et al., 2008). The deficiencies of guidelines are
confirmed in the pilot study by the fact that the information given is not sufficient
to the designers’ needs and they have to supplement it with tests and evaluations.
152 Zitkus et al.
15.3.2 User Participation in the Design Process

Although some of the interviewees mentioned that they could test the features of a
new concept design with potential users, the majority of the comments indicate that
users are not part of the design routine. Normally, as described in the interviews,
the design activity does not include elderly or disabled people in user trials, even in
projects where they are the target market. Designers rely on their own assessments,
limited as they are by not being able to feel other individuals’ capabilities.
In addition, accessibility and usability are described as only some among the
numerous requirements that designers have to comply with. User requirement is
not a special requirement above the others, which means that designers have to
balance the user’s needs with functionality, manufacturing, maintenance,
environmental aspects, aesthetics and other requests. Moreover, it was also
highlighted that accessibility does not always have a place in the product design
requirements.
The deficiencies in the mechanisms that provide user information to designers,
added to the fact that user-trials are rarely adopted as part of the design process,
result in the need already mentioned at the beginning of the chapter: the need to
integrate user requirements into the design activity.
15.3.3 How Interactive Interfaces Provide Information

Two aspects of interactive techniques were highlighted in the interviews after the
demonstration; one is the interface, and the other is the information provided.
15.3.3.1 The Usability of the Interface

The outcomes of the demonstration of the interactive technique proposed in this
study confirmed an aspect already highlighted by Macdonald et al. (2006). The
authors highlighted that the integration of supportive techniques with the
designer’s tool is a way to enhance the acceptability of the technique. In the study
carried out by Macdonald et al, the integration was with CAD models, whereas the
designers’ responses included other graphic tools as well.
Although some designers stressed that an interactive tool should be built into
(or within) CAD, others underlined that the technique should provide the
information according to the design practice and its related tool. There are reasons
for both opinions. On the one hand, industrial designers tend to design in CAD and at
some point all the information about the new concept, including graphic information,
will be there for photorealistic renderings, which explains the easy integration with
CAD. On the other hand, there are some variations across different design domains.
For instance, according to this study, packaging design does not necessarily use CAD
software. Although the design generally goes to computer graphics (Corel Draw for
instance) before going to the laser cut, there is no need of a 3D digital model as the
drawing has to be the opened/flat cut-out of the cardboard. Moreover, all the graphic
design (the visual communication) to feature on the package does not need to get into
CAD either: it can be done in Photoshop or Illustrator, both 2D graphic software.
The above findings stress the need for generating additional interactive interfaces
to be integrated and tested in a range of computer graphic tools. This would test the
usability of the techniques by getting more feedback from the intended users, in this
case the designers.
15.3.3.2 The Information Provided in the Interface

It seems that normally designers think about a target market from which a reasonable
amount of people from the entire population are excluded. Therefore, unless clients
request or designers understand the benefits of changing the design, the objective of
providing information about inclusion will not succeed.
Information on inclusion, however, is not only a matter for designers; as already
mentioned by Gill (2009) small to medium-sized design consultancies tend to face
the pressure of costs and tight deadlines from the client, which constrain the
designer’s decisions. Among all the interviewees the role of the client was always
underlined as the main supplier of information, which includes information on
potential consumers and their needs. Moreover, if the client does not act as an
information supplier, they guide the designers to the kind of information they have to
get through research. Therefore, further investigation is needed to understand which
type of inclusive information that supports the design activity could also be practical
for the client.
15.4 Conclusions
Computer graphic systems seem to be well integrated in the design activity, and thus
they are an important means of supplying designers with knowledge about inclusion.
This knowledge about inclusion is necessary as designing is a trade-off activity that
commonly does not prioritise accessibility among other requirements. However, to
implement this scenario the techniques have to consider not only the design activity,
but also the client’s view about the inclusive information supplied. Therefore, further
research has to be carried out to understand how the benefits of knowledge about
inclusion could be explicit to both designers and clients.
The authors would like to thank Dr Alaster Yoxall and Prof Steve Gill for their
support of this study, and also all the designers involved.
15.6 References
Burns CM, Vicente KJ, Christoffersen K, Pawlak WS (1997) Towards viable, useful and usable
human factors design guidance. Applied Ergonomics, 28(5-6): 311-322
154 Zitkus et al.
Cardoso C, Keates S, Clarkson PJ (2003) Assessment for inclusive design. In: Clarkson PJ,
Coleman R, Keates S, Lebbon C (eds.) Inclusive design: Design for the whole population.
Springer-Verlag, London, UK
Cardoso C, Clarkson PJ (2007) User simulation in product evaluation. In: Coleman R, Clarkson
PJ, Dong H and Cassim J (eds.) Design for inclusivity: A practical guide to accessible,
innovative and user-centred design. Gower Publishing Ltd, Hampshire, UK
Clarkson PJ, Coleman R, Hosking I, Waller SD (2007) Inclusive Design Toolkit. Engineering
Gill S (2009) Six challenges facing user-oriented industrial design. The Design Journal, 12(1): pp
41-67
Goodman-Deane J, Langdon PM, Clarkson PJ (2010) Key influences on the user-centred design
process. Journal of Engineering Design, 21(2-3): pp 345-373
Hitchcock D, Taylor A (2003) Simulation for inclusion - true user centred design. In: Proceedings
of the International Conference on Inclusive Design and Communications (INCLUDE 2003),
London, UK
Law CM, Yi JS, Choi YS, Jacko JA (2008) A systematic examination of universal design
resources: Part 1, heuristic evaluation. Universal Access in the Information Society, 7(1-2): 31-
54
Macdonald AS, Loudon D, Rowe PJ, Samuel D, Hood V, Nicol AC, et al. (2006) InclusiveCAD:
A software resource for designers. In: Clarkson PJ, Langdon PM, Robinson P (eds.) Designing
accessible technology. Springer, London, UK
Marshall R, Case K, Porter JM, Sims R, Gyi DE (2004) Using HADRIAN for eliciting virtual
user feedback in ‘design for all’. In: Proceedings of the Institution of Mechanical Engineers,
Part B: Journal of Engineering Manufacture, 218(9): 1203-1210
Newell AF, Carmichael A, Morgan M, Dickinson A (2006) The use of theatre in requirements
gathering and usability studies. Interacting with Computers, 18(5): 996-1011
Norman DA (2002) The design of everyday things. Basic Books, London, UK
Rode JA, Toye EF, Blackwell AF (2004) The fuzzy felt ethnography: Understanding the
programming patterns of domestic appliances. Personal and Ubiquitous Computing, 8(3-4):
161-176
Sanders EB-N (2000) Generative tools for co-designing. In: Scrivener SAR, Ball LJ, Woodcock
A (eds.) Collaborative Design, Springer-Verlag, London, UK
VERITAS D4.1.3_v2 (2010) Project presentation and project description leaflet. Fraunhofer IAO,
Stuttgart, Germany
Zitkus E, Langdon PM, Clarkson PJ (2011) Accessibility evaluation: Assistive tools for design
activity in product development. In: Proceedings of the 1st International Conference on
Sustainable Intelligent Manufacturing, Leiria, Portugal
Zitkus E, Langdon PM, Clarkson PJ (2012) How are user requirements dealt with in the design
process. In: Proceedings of the 12th International Design Conference (DESIGN 2012),
Dubrovnik, Croatia, in press
Chapter 16
From Guinea Pigs to Design Partners:

Working with Older People in ICT Design
R. Edlin-White, S. Cobb, A. Floyde, S. Lewthwaite,
J. Wang and J. Riedel
16.1 Introduction
16.1.1 Ageing Societies and the Implications for Research
Governments worldwide are recognising the reality and challenges of aging
societies. Aging is often accompanied by increased incidence of impairments in the
physio-motor, sensory and cognitive domains, as well as health problems, reduced
socialisation, poorer finances, reduced sense of purpose and sometimes
marginalisation from society. Technology can be very beneficial for older people
but is too often problematical, sometimes creating digital exclusion.
Researchers and funding bodies are beginning to focus specifically on the ICT
needs of older people, and excellent work is being conducted, but as yet research
methods and models of engagement for working with older people are fragmentary
or at an early stage of development (Rice and Carmichael, 2007).
This paper explores methods and approaches currently used in HCI research
with older users, and those used by the authors in the MyUI European research
project, and presents a list of themes which the authors believe will be relevant to
other work in the area and could inform a more consistent methodology.
16.1.2 User-centred Methods and Approaches in HCI

A critical problem in the design of usable, accessible technology is that most
designers are highly motivated by technology per se and have an inadequate
understanding of the world of end users. Their effectiveness at user-centred design
may be impaired by positivist world views, assumptions that user experiences are
similar to their own, and a propensity for confirmation bias when determining user

156 Edlin-White et al.
needs (Bader and Nyce, 1998; Norman, 1998; Simonsen and Kensing, 1998).
Design teams and processes should be more user-centred.
Muller et al. (1997) noted the chasm between the worlds of designers and users,
each world having its own culture, language, space and assumptions. The HCI
community, they say, may need to develop “anthropological and sociological
methods for working across differences in language and abilities” to improve “the
field’s competence in intercultural communications”. This is even more necessary
when users are older, retired people, and designers are young and employed.
Cultivating a developing rapport with user communities, through field study, or
better still by including users in the design process, can help designers bridge the
chasm, understand the world, work and requirements of users, and by combining
User Needs Analysis and Requirements Engineering can lead to more accessible
and effective products (Muller, 2002; Lindgaard et al., 2006).
Muller presents a thorough, persuasive account of Participatory Design. User
engagement, he finds, is too often one-directional, creating applications (of
technology) rather than solutions (to user problems). He appeals for mutual
learning and openness to surprise (Muller, 2002; see also Farrel et al., 2006).
Crabtree (1998) believes ethnography can usefully complement Participatory
Design in user studies. People cannot always describe their own work effectively,
but an observer can see and record the activity. An approach based on reviews of
prototypes, he says, rarely empowers the users to challenge underlying premises
such as “are we solving the right problem?”
16.1.3 HCI Methods Currently Used with Older Participants

It has been argued that standard HCI approaches to user engagement are not
appropriate when working with older people (Eisma et al., 2003, 2004; Rice and
Alm, 2008). Much academic HCI research has used (mainly young) students as the
main source of participants, age-biasing not only research findings but also the
evolution of research approaches and methods. Older people, and those with
disabilities, are currently under-represented as participants in academic research,
especially HCI research (Lewis, 2006; Newell et al., 2007; Walker, 2007).
Some HCI researchers working with older users have published their methods,
usually in a descriptive way, not formulated as a standard recommended approach.
Newell and Gregor (2000) discuss “User Sensitive Inclusive Design” - a
paradigm for including users with disabilities, including older users, within a
development framework. They appeal for respectful user involvement, but also
note that - for a variety of reasons - potential end-users are not always the sole or
best source of information about how a novel system may be of benefit to them or
their diverse peers. They and other researchers (Muller, 2002; Hansson et al., 2006;
Rice and Carmichael, 2007) recognise that other parties - partners, carers,
clinicians or other professionals, teachers, family members and union
representatives - can provide valuable supplementary insights.
Eisma et al. (2003) appeal for a process involving “mutual inspiration” when
working with older people; this can happen most readily in informal studies and
interactions with the participants; in enjoyable, sociable group activities with
From Guinea Pigs to Design Partners 157
refreshments, and engaging in hands-on activities using research-related stimuli.

This can provide rich, grounded and often serendipitous findings. A questionnaire
conducted in a group context can stimulate conversations which generate more
findings than the written responses. Observation by researchers can reveal more
information than user reports alone (Eisma et al., 2003).
Working with older and disabled people as users, Picking et al. (2009)
developed the EDUCATID framework, with ethical considerations and user
engagement at the heart of the process. The ethical approach is based on Abascal
and Nicolle’s (2005) detailed and thoughtful discussion of ethics in HCI work. A
genuine and open concern for the user can highlight impacts and consequences of a
technological intervention which might otherwise be overlooked. In EDUCATID,
user narratives are elicited through workshops, focus groups etc. and these are used
to generate requirements and later concept designs. The method continues with
formative and summative evaluations of prototypes and systems. Focus Groups are
noted as useful in this field by many researchers (Barrett and Kirk, 2000; Eisma et
al., 2003; Goodman et al., 2004).
The OASIS project (Lindsay et al., 2008) developed the Oasis Participatory
Analysis Framework (OPAF), for involving older users throughout a design
process, including some very practical guidance in areas like recruitment. They
advocate group discussions, focus groups and design workshops in field settings as
methods of engagement, which they believe are less time consuming than pure
ethnographic methods. They also commend extensive use of PICTIVE techniques
(Muller, 2002) and low fidelity prototyping, in workshop and group sessions.
Similar advice to all the above, most relevant for plentifully resourced work, is
provided by the respected CREATE centre for research on aging and technology in
chapters 3, 12 and 13 of Designing for Older Adults (Fisk et al., 2009).
Few quantitative methods appear to be used. Pernice and Neilsen (2002)
describe a number of qualitative and quantitative studies with older participants,
concluding (p122) that quantitative studies need a large number of participants to
achieve statistical significance, are time consuming to design and set up, and
generate good numeric data but fewer insights than qualitative methods.
Qualitative methods, while generating less objective data, and with the
generalisability of the findings sometimes open to question, are better at creating
deep and rich understanding, and are therefore particularly well suited to
generating ideas and hypotheses, in understanding fields characterised by
ambiguity and complexity, and in requirements elicitation and early evaluations in
a process of iterative refinement (Creswell, 2007; Lamsweerde, 2009).
The above and other related literature informed the approaches adopted in the
MyUI project. It provided valuable insights but leaves many possibilities open.
16.2 Involvement of Older People in MyUI Project

16.2.1 The Project
MyUI is a European Union funded research project which sets out to facilitate the
development of highly accessible life-enhancing technology-mediated services for
older people, including those who are experiencing age-related changes and
impairments in the physio-motor, sensory or cognitive domains. Web-based
services (in areas such as social communication, simple exercise programmes
monitored and modified by experts, entertainment, education and reminders) will
be delivered through the attractive medium of interactive TV using traditional or
novel hand-held remote controls. Accessibility will be enhanced by the use of
adaptive systems which detect a user’s changing accessibility needs during normal
usage and adapt the user interface in real time to better match those needs.
The University of Nottingham Human Factors Research Group was responsible
for establishing the methodological approach, coordinating all end-user studies
across Europe and conducting most of them.
16.2.2 Methodology for Participant Involvement

It is clear that most of the main objectives of MyUI - to build accessible, life-
enhancing, valued services for older people - are based on subjective social
constructs. Notions such as “beneficial”, “accessible” and “usable” will vary from
person to person. A quality of life measure must include the viewpoint of the
person living that life. People who make an informed choice to opt out of a
technology are not necessarily “lost sheep” (Klecun, 2008). This understanding
profoundly affects our style of engagement with communities of older people. We
aim to listen to their articulations of these realities rather than imposing our own.
Based on experience of previous projects and on the literature reviewed above,
a hybridised and evolving methodology has been adopted, influenced by
Participatory Design, ethnography and Participatory Research, and shaped by
social constructivist models. Participants are treated as co-creators, the foremost
experts in their own needs, whose contribution is recognised and valued as a vital
part of the work. In Radermacher’s (2006) classification of participatory methods
our main approach is “Researcher initiated/shared decisions with participants”.
The project has also involved some empirical studies with older people, an
approach based on a more positivist paradigm, in which the older person is - at
least temporarily - a subject (or “guinea pig”) rather than an equal participant. In
order to avoid Hawthorne effects, the purpose of the study and the nature of
researcher observations were sometimes not fully disclosed until the end of the
study, and the researchers temporarily adopted a slightly more detached stance.
This was prompted by demands for “scientific” studies, and has provided
convincing, formative findings, but as an approach sits somewhat uneasily with our
ethical stance and quality of relationships with our user communities.
16.2.3 Ethical Approach

Our ethical approach on this project goes well beyond meeting bureaucratic
requirements. An ethical stance should permeate every aspect of the work, treating
all stakeholders - but especially any who are vulnerable - with honesty, respect,
courtesy and care. We believe the research should benefit the participants or future
generations of people in similar circumstances, and not just our own interests as
researchers. Informed consent is “gently invited”, recognising that hearing,
reading, remembering etc. may be impaired. If later a participant asks us what
we’re doing or why, we willingly reiterate and remind them they can withdraw.
Part of our stance is the belief that accessibility problems generally arise from
poor design, not inadequate users, so we avoid the use of pejorative terms such as
“computer literacy” and are more concerned to ensure our designs should be
informed by “user literacy”. Where possible we avoid medical questions and
instead ask accessibility questions; e.g. we ask “would this display be adequately
readable for you”, and not “is your eyesight adequate for this display”. Our
questionnaires and consent forms are designed so they can also be administered
verbally, in case the participant has any problems with reading or writing, e.g. due
to literacy, eyesight, or digital dexterity limitations.
We attempt to use age-appropriate measures of accessibility, usability or
benefit of a technology, as older and retired people tend to have different priorities.
These measures are mainly subjective, e.g. enjoyment, satisfaction, feeling of
autonomy, mastery, ease of use and ease of understanding, and not feeling
patronised or frustrated by the technology.
We recognise that placing a technological artefact into someone’s domestic
environment and expecting them to use it can have far-reaching effects in many
domains. For example, enabling someone to do online shopping might affect the
frequency of going out to shop, and hence their amount of socialisation, exercise,
sunlight and fresh air, perhaps therefore affecting their physical and mental health.
It might reduce the amount of help with shopping from family or neighbours -
again reducing social contact. Different grocery buying patterns may affect the diet
and finances for better or for worse. Provision of technology may lead to a need for
training, support and maintenance activities which may or may not be satisfied.
Some ICT projects for older users have concluded with brief and superficial
usability studies, optimistically interpreted, failing to take into account the benefit
of support and training from researchers which could never be afforded in a
commercial product (Dickinson and Gregor, 2006; Blaschke et al., 2009). We
believe it is ethically appropriate to also try to estimate the broader Quality of Life
(QoL) so that these many factors are not ignored.
16.2.4 Study Settings

We established regular contact with three centres where older people routinely
congregate: two Day Centres (DC1, DC2) and a self-contained retirement village
(RV1), between them representing over 600 older people and providing a good
range of contrasting demographics.
Through our Spanish consortium partners we organised research in an older
people’s complex in Getafe, near Madrid. There were clear contrasts in lifestyle
and aging and therefore in requirements. Compared to our UK groups, the Getafe
cohort’s lifestyles seemed to include more swimming, dancing, and sociable
outdoor activities (perhaps due to the climate) but less literacy. A popular activity
among the Getafe women was making costumes for fiestas and carnivals (Edlin-
White et al., 2010). Older people with such physically, socially and cognitively
active lifestyles are less susceptible to age-related impairments.
All research activities have taken place in field settings - DC1, DC2 and RV1.
Lab-based studies with older people can create artificial responses due to changes
to their routine, as well as creating practical problems (e.g. transport, diets, carers
etc.) and sometimes duty of care issues (Edlin-White et al., 2011). However studies
carried out in “open” field settings, e.g. a day room, can be susceptible to all
manner of distractions. During one of our early focus groups, staff brought round
trays of sherry to mark a 40th wedding anniversary. It seems uncharitable to
describe this as a confounding factor and not ecological validity.
To improve research focus, we created unobtrusive “pop-up” labs in these
locations, by arranging to use a side room and bringing in sufficient equipment and
materials to create a mini usability lab. This allowed fewer distractions, more
experimental controls, a more formal and comparable participant experience, and
in individual studies eliminated the tendency for participants to influence each
other’s responses. For some studies, within the implanted lab we created an inner
environment (furniture, décor, lighting and equipment) to emulate a domestic TV
watching experience for the participant.
16.2.5 Study Methods - Progressive Engagement Process

The research involved significant journeys of increasing trust and collaboration for
researchers, staff and participants.
Initial contact was through senior staff (“gatekeepers”) at each venue. Early
communication included presentations at a non-academic level, to explain the
project and our research aspirations, stressing aspects likely to be of interest to this
audience. In all cases the result has been positive; staff believed that older people
are not consulted sufficiently in technology design.
Initially study design was heavily influenced by the need to build trust and to
ensure staff and participants believed the visit was enjoyable and beneficial. This
meant working within the unique established culture of each venue and user
community. In DC2, for instance, this meant working with a group of around 18
older people with a variety of abilities and limitations, sitting on seats arranged in a
horseshoe shape along three walls of a large room. The research opportunity might
be a 45 minute slot following a raffle and preceding a seated exercise session, and
there would be the sound of a radio playing light classical music from the adjacent
kitchen as well as the smells of food being prepared. Within this context it was
possible for us to address the whole group, to obtain consent, and to conduct an
informal focus group, stimulated by handing round a variety of hand-held devices

(iPod, phone, TV remote, DVD remote, calculators etc.). In DC1 people were
often sitting “cabaret style” around small tables in groups of 4 to 6, perhaps
knitting, chatting, doing puzzles, etc.. There would be periodic activities led from
the front using a microphone, e.g. quizzes. We therefore used this “pub quiz”
format to conduct early research.
As trust and rapport grew, it became possible to increase the research rigour
and shape the culture accordingly; e.g. quiz or puzzle formats but including
questions about technology attitudes and aptitudes. With staff support it became
possible to create the pop-up lab implants, and invite participants to leave the
communal area for short “activities” (a better word than “studies” for this
audience). In early visits we avoided asking personal questions, e.g. concerning
disabilities; latterly it became acceptable to include such questions. The responses
help to interpret study results and to recruit more purposefully. Consent for
photography, video and audio recording was forthcoming too, as trust grew.
Research methods have included focus groups, collaborative participatory design
workshops, questionnaires, interviews, usability studies and empirical studies.
Participant selection within the venues was initially opportunistic, based on
availability and willingness of participants and the possibilities afforded by
communal space or pop-up labs. Later, more purposeful sampling was used.
Recruitment at RV1 was initially by advertising an informational presentation
event. Further recruitment has been through networking, involvement and
announcement at their well supported Monday night pub quiz, and with the
assistance of two enthusiastic residents who have acted as recruiting champions.
We have not offered financial incentives to participants; staff advised these
would be unnecessary, but refreshments (hot drinks and cakes) have been
provided, especially for longer (>45 minute) sessions, or to facilitate group
interaction. We have instead explored whether we can support the venue, e.g. by
provision of equipment or expertise. Most participants seem glad to be involved,
embracing a volunteering ethos, hoping that it may lead to better technology in the
future, even if not in their lifetimes. Significant effort has been dedicated to
keeping the staff and participants at all three locations informed, agreeable and
supportive. We strive to make the research activities enjoyable for participants.
Afterwards we seek feedback from staff to ensure they are happy with our visit and
with the well-being and enjoyment of the participants during the study.
Participants have taken on a number of roles. In our few empirical studies, they
have to some extent been subjects (or “guinea pigs”) rather than participants.
However, in most activities the participant is engaged as an expert in their own
accessibility needs and technology requirements, and indeed a valued design
partner. Some of the more able participants have also acted as proxies for others
not in the group, perhaps with greater accessibility needs. Some of our questions
target this proxy role, with phrases such as “you or other older people”. More able
participants sometimes act as carer or support for a less able person; e.g. helping
them with understanding or reading or writing. Some of our more enthusiastic and
supportive older people have become champions for our cause, helping with
recruitment, room booking and even in aspects of facilitation of sessions.
16.3 Conclusions
It appears that there is significant expertise but limited methodological guidance in
this field. Methods and approaches are evolving. The MyUI project drew on the
literature but also learned much through experience and continuous refinement of
methods. We have identified the following common themes from both literature
and project experience; themes which would have been beneficial to us at the start
of our project, may be relevant for future projects, and may contribute towards
more formal methodological recommendations.
• Recognise that HCI work with older people is social research, based on
subjective social constructs such as accessibility, usability and quality of
life.
• Adopt an ethical and user-centred approach, perhaps utilising elements of
Participatory Design, Accessible Design and Ethnography.
• User perspective to be taken seriously and respectfully, but self-knowledge
can be incomplete or imperfect. Include supplementary perspectives.
• Allow time in the project plan to build trust with participant groups and
gatekeepers, and cultivate good working relationships on an ongoing basis.
• Study settings - mainly field settings for ecological validity. Pop-up labs
implanted in field settings can be effective, though time consuming.
• Recognise the challenge and difficulty of recruiting a representative
sample.
• Study methods - Focus Groups and Design Workshops are effective; also
questionnaires and interviews and possibly retrospective verbal protocols.
• Design all aspects of studies to accommodate participants with very varied
abilities. This includes the Informed Consent process. Allow for different
learning speeds and varied learning styles of older people.
• Quantitative methods with inferential statistics are unlikely to be effective
unless there is a lot of time and resources and access to many participants.
• Study measures need to be appropriate to older users, and will probably
include more subjective measures than mainstream HCI studies.
• Be flexible and open to change while conducting a study. Allow time for
“social niceties” and off-topic digressions - some of which prove to be
useful.
• Quality of life impacts of technology are important. QoL impact
measurement is difficult, costly and most properly conducted over a long
time period.
It is hoped that our experiences documented here, together with those of others,
will contribute towards the development of more fully formed and finely honed
methodological guidance for involving older people in technology design.
16.4 References
Abascal J, Nicolle C (2005) Moving towards inclusive design guidelines for socially and
ethically aware HCI. Interacting with Computers, 17(5): 484-505
Bader G, Nyce JM (1998) When only the self is real: Theory and practice in the
development community. Journal of Computer Documentation, 22(1): 5-10
Barrett J, Kirk S (2000) Running focus groups with elderly and disabled elderly participants.
Applied Ergonomics, 31: 621-629
Blaschke C, Freddolino P, Mullen E (2009) Ageing and technology: A review of the
research literature. British Journal of Social Work, 39(4): 641-656
Crabtree A (1998) Ethnography in participatory design. In: Proceedings of the 1998
Participatory Design Conference, Seattle, WA, US
Creswell J (2007) Qualitative inquiry and research design: Choosing among five
approaches. Sage, Newbury, CA, US
Dickinson A, Gregor P (2006) Computer use has no demonstrated impact on the well-being
of older adults. International Journal of Human-Computer Studies, 64(8): 744-753
Edlin-White RW, Floyde A, D’Cruz M, Cobb S, Riedel JCKH (2010) User interface
adaptation for accessibility by older users and the challenges of cultural diversity. In:
Proceedings of the 14th International Conference on Human-computer Interaction:
Towards Mobile and Intelligent Interaction Environments (HCII 2011), Orlando, FL, US
Edlin-White RW, Cobb S, D’Cruz M, Floyde A, Lewthwaite S, Riedel JCKH (2011)
Accessibility for older users through adaptive interfaces: Opportunities, challenges and
achievements. In: Proceedings of the 14th International Conference on Human-computer
Interaction: Towards Mobile and Intelligent Interaction Environments (HCII 2011),
Orlando, FL, US
Eisma R, Dickinson A, Goodman J, Mival O, Syme A, Tiwari L (2003) Mutual inspiration
in the development of new technology for older people. In: Proceedings of the
London, UK
Eisma R, Dickinson A, Goodman J, Syme A, Tiwari L, Newell AF (2004) Early user
involvement in the development of information technology related products for older
people. Universal Access in the Information Society, 3(2): 131-140
Farrell V, Farrell G, Mouzakis K, Pilgrim C, Byrt P (2006). PICTIOL: A case study in
participatory design. In: Proceedings of the 18th Australia conference on Computer-
Human-Interaction (OZCHI ’06), Sydney, Australia
Fisk AD, Rogers WA, Charness N, Czaja SJ, Sharit J (2009) Designing for older adults:
Principles and creative human factors approaches. CRC Press, London, UK
Goodman J, Dickinson A, Syme A (2004) Gathering requirements for mobile devices using
focus groups with older people. In: Keates S, Clarkson PJ, Langdon P, Robinson P (eds.)
Designing a more inclusive world. Springer, London, UK
Hansson C, Dittrich Y, Randall D (2006) How to include users in the development of off-
the-shelf software: A case for complementing participatory design with Agile
development. In: Proceedings of the 39th Annual International Conference on System
Sciences, Kauai, Hawaii, US
Klecun E (2008) Bringing lost sheep into the fold: Questioning the discourse of the digital
divide. Information, Technology and People, 21(3): 267-282
Lewis C (2006) HCI and cognitive disabilities. Interactions, 13(3): 14-15
Lindgaard G, Dillon R, Trbovich P, White R, Fernandes G, Lundahl S et al. (2006) User
needs analysis and requirements engineering: Theory and practice. Interacting with
Computers, 18(1): 47-70
Lindsay S, Jackson D, Olivier P (2008) Guidelines on UCD Requirements extraction.

Available at: http://irgen.ncl.ac.uk/oasis/D5.5.2.doc (Accessed 4 September 2011)
Muller MJ (2002) Participatory design: The third space in HCI. In: Jacko JA, Sears A (eds.)
The human computer interaction handbook: Fundamentals, evolving technologies and
emerging applications. Lawrence Erlbaum, Mahwah, NJ, USA
Muller MJ, Wharton C, McIver WJ, Laux L (1997) Toward an HCI research and practice
agenda based on human needs and social responsibility. In: Proceedings of CHI 97
Workshop on Human Factors in Computer Systems, SIGCHI, Atlanta, GA, US
Newell AF, Gregor P (2000) User sensitive inclusive design - in search of a new paradigm.
In: Proceedings of the ACM Conference on Universal Usability (CUU’00), Washington,
DC, US
Newell A, Carmichael J, Morgan M (2007) Methodologies for involving older adults in the
design process. In: Proceedings of the 4th International Conference on Universal Access in
HCI, Beijing, China
Norman DA (1998) The invisible computer. MIT Press, Cambridge, MA, US
Pernice K, Nielsen J (2002) Web usability for senior citizens. Design guidelines based on
usability studies with people age 65 and older. Nielsen Norman Group, Fremont, CA, US
Picking R, Robinet A, Grout V, McGinn J, Roy A, Ellis S et al. (2009) A case study using a
methodological approach to developing user interfaces for elderly and disabled people.
The Computer Journal, 53(6): 842-859
Radermacher HL (2006) Participatory action research with people with disabilities:
Exploring experiences of participation. PhD Thesis, Victoria University, Victoria,
Australia. Available at: http://wallaby.vu.edu.au/adt-VVUT/uploads/approved/adt-
VVUT20060904.160428/public/02whole.pdf (Accessed 10 March 2008)
Rice M, Carmichael A (2007) Effective requirements gathering for older adults. Computing
and Accessibility Special Issue on Accessible Europe, SIGACCESS Accessible
Computing Newsletter, 88: 15-18
Rice M, Alm N (2008) Designing new interfaces for digital interactive television usable by
older adults. ACM Computers in Entertainment, 6(1)
Simonsen J, Kensing F (1998) Make room for ethnography in design! The Journal of
Computer Documentation, 22(1): 20-30
Van Lamsweerde A (2009) Requirements engineering: From system goals to UML models
to software specifications. John Wiley & Sons, London, UK
Walker A (2007) Why involve older people in research? Age and Ageing, 36(5): 481-483
Chapter 17
When Users Cannot be Included in

Inclusive Design
R. Herriott
17.1 Introduction
Inclusive Design (ID) methods place a strong emphasis on user participation in
designing mainstream products. In recent years researchers in the field of assistive
technology (AT) have drawn on and contributed to the ID approach. There are
good grounds for this association. However, the linkage elides the differences in
methods that are available and appropriate to designers in the respective fields. The
demands made by strategies such as co-creation, focus groups, cultural probes and
even simple interviewing can be above the capacities of the users of AT. Yet the
impairments of ill and disabled users make the need for usability and pleasurability
(Green and Jordan, 2002) even more important since alternative products are
comparatively few. This paper examines the workarounds two teams of designers
have used to reduce the demands placed on emphysema patients and elderly users
during inclusive design processes. In the case of a student design project it was
necessary to focus on a super-user, use prototyping as a creative tool and to use
improvised ergonomic simulation. In a second case a consultancy was required to
place more emphasis on ethnographic, observational methods and personas where
co-creation and co-design proved to be beyond the capacities of the user-group.
17.1.1 Structure
The paper begins with a discussion of ID and its relation to AT design. There is an
overview of approaches to AT. This is followed by a description of two cases. The
first concerns a breathing apparatus for patients with chronic obstructive
pulmonary disorder. The second is a municipally funded research project into
innovations to help the elderly and handicapped. Both projects were based in
Denmark. To conclude there is a discussion of the findings and their implications.

166 Herriott
17.2 Inclusive Design and AT

The literature on inclusive design shows a close linkage between it and AT. The
CWUAAT series (2004 onwards), and the Include Conference series (2005 onwards),
show a strong association between the methodology of inclusive design and the goals
of AT products (e.g. Orpwood et al., 2004, Dhiensa et al., 2005, Mayagoitia et al.,
2006, Mountain et al., 2006, Orpwood et al., 2008, Linnott, 2011). There is an overlap
of interests driving this association. Inclusive design is a set of methods to optimise
product designs so as to accommodate a much broader range of capabilities (BSI,
2005). AT in contrast is not a design method but a set of solutions to the problems
caused by chronic ill-health or long-term disability. There are two ways to look at the
distinction. One is that AT products have an element of compulsion lacking in
mainstream consumer goods. If one wants to walk despite having weakened leg
muscles, one must use a walking stick or rolling-frame. In contrast, it is only in a broad
sense that one is compelled to own many consumer goods. Such goods are additions to
our range of standard capabilities rather than replacements. AT products stand in where
patients’ abilities have declined severely or were not there originally: the wheelchair,
hearing aid and speech synthesiser are not products the average user will seek to
purchase. Furthermore, potential AT users may even prefer to avoid using such
products even if they could help; this is sometimes the case because the stigmatising
appearance of the product is considered worse than the problems caused by not using it
(e.g. Bichard et al., 2007). Another way to consider the cleavage between mainstream
ID and AT is that the latter represents the point at which there is a step-change in the
needs and capabilities of the user groups. A person who can read a newspaper with
glasses in average light is qualitatively different from the person who effectively cannot
read standard text. A person who walks comfortably though slowly is qualitatively
different from one for whom walking demands 100% of their respiratory capability. It
is worth pointing out that the binary distinction between impaired and mainstream users
is a somewhat artificial construct since ability loss can vary in many cases. However, in
the cases discussed in this paper there were many indicators which could objectively
point towards the users being clearly on that side of the continuum where AT devices
were the appropriate design solution.
Inclusive design is thus a generally applicable range of tools and strategies
originally devised to improve mainstream product design. These tools are aimed at
making users a bigger part of the design process resulting in more valid design
solutions (e.g. Coleman et al., 2007). While assistive technology can take inspiration
from the ideas of better design and higher quality aesthetics embodied by inclusive
design, it must be recognised that the user groups have qualitatively different
characters. The design processes for each therefore can not be the same. I am not here
suggesting that AT designers think that their users´ profiles are identical with the
profile of the broader average. I wish instead to draw attention to the need for a
different emphasis in using the tools devised from one area, mainstream design, in
another area, assistive technology. The paper proceeds from this starting point to
examine how designers manage the different capabilities of users who are unable to be
as large a part of the design process as designers might wish.
When Users Cannot be Included in Inclusive Design 167
17.2.1 Approaches to AT Design

One approach taken is to view AT as a branch of medical technology which uses
design methods allied to those used for medical and rehabilitation devices (Cook
and Hussey, 2002). This has tended to use a hard systems model of development:
“Within design for disability, where teams tend to come from clinical and
engineering backgrounds, the dominant culture is still one of solving problems”
(Pullin, 2009). A second approach is to borrow from the more user-centred, social-
science based methods of ID. As Reed and Monk (2006) put it: “The key to
understanding user-centred design is the aphorism ‘Know your user.’” This is a
more recent development. The research presented at CWUAAT and at RCA’s
INCLUDE conferences shows that this second approach is still a developing field.
Torrens (2010) classifies the two approaches as being respectively quantitative and
qualitative. Torrens suggests a third, synthetic approach using mixed methods
which “offers the optimum opportunity for triangulation: validation of data through
multivariate methods of collection” (ibid.). This synthetic approach still assumes -
not unreasonably - a significant role for the user in the design process.
The three approaches to AT differ in the extent to which social-science
methodology is deployed. Whichever approach is chosen, the range of tools to
bring the user into the design process is the same: questionnaires, video
ethnography, interviews, focus groups, critical user forums, cultural probes, and
workshops. Ideally, the user is also consulted when the information initially
gathered is turned into a specification and again after concept designs are translated
into alpha and beta prototypes (Clarkson et al., 2007).
Ideally, ID methods should be as comprehensive as possible. This builds
redundancy into the information gathering process so that what one method may
miss another may pick up. But such redundancy can be extremely demanding for
patients with debilitating conditions such as emphysema, severe arthritis or
muscular dystrophy. This situation points to an asymmetry in the inclusive design
process. While the aim of the process is to design a more inclusive product, the
input process of information gathering, processing and validation may itself be
violating the sixth principle of universal design: low physical demand (Story et al.,
1998).
17.3 Cases
There now follow two cases of how designers responded to the diminished
possibility for user-involvement in the design of AT products. The information was
derived from interviews carried out by the author. The interviews were recorded
and transcribed. For clarity the quotes have been edited.
168 Herriott
17.3.1 Breathe Easy

Tonirie Sembach Lauridsen and Marco Barbiani, students at the Aarhus School of
Architecture, identified problems with the existing oxygen supply systems for
patients with reduced lung-capacity. The project was carried out in spring 2011.
Their user-group consisted of patients with chronic obstructive pulmonary disease
(COPD). This is “a major cause of morbidity and mortality. In Europe COPD,
together with asthma and pneumonia, is the third most common cause of death. It
is a disorder characterised by reduced maximum expiratory flow and slow forced
emptying of the lungs, features which do not change markedly over several
months” (Jeffery, 1998). This condition results in a sharply reduced capacity for
physical activity and a much reduced will for and interest in non-essential tasks.
COPD patients supplement their air intake using wearable inhalers. The limited
physical capacity of these patients requires that the equipment is as a light as
possible and that it requires the least physical effort to operate. As a product for
people with severe lung disease, Breathe Easy falls into the category of assistive
technology. It demonstrates the added difficulties of working with ill users and
keeping them involved as much as is required by the ideal inclusive design process.
Figure 17.1. Breathe Easy portable unit
Initially, the designers formulated a design strategy approximately conforming

to that recommended by the Cambridge Engineering Design Centre (EDC).
However, they discovered that their users were unable to participate very fully in
the design process due to their medical condition. The planned design process was
one where the users played a key role as per ID best practice. When asked if the
design process followed the planned design process the answer was “No.
Definitely not” (Barbiani, 2011).
The format of Barbiani and Lauridsen’s design was that of the concentrator,
which extracts oxygen from the atmosphere. This class of device has a
concentrator which is a large stationary element (principally a tank) and a smaller
portable reserve carried by the patient. Technical problems related to the reloading
of the portable unit, making the portable part comfortable to wear and managing
the inconvenience of the nasal catheter. The existing designs required more force
to connect the portable unit to the concentrator than the users were comfortable
with so a simpler connection was needed. This design also had to accommodate
the difficulty of treating the super-cooled connecting elements. In existing products
these are exposed thus presenting the risk of frost-burn. Finally, the concept had to
look more like a regular consumer good than something medico-industrial. Users
resented having scuffed, crude-looking equipment in their homes.
Ideally, inclusive design involves user-centred design methods in order to bring
users actively into the design process. The students discovered that a combination
of sick users and a technically demanding product rendered these objectives
difficult to attain. The result was a project which aimed for user-centred design but
where the users were few in number and often too sick to interact with. When
asked whether the designers had a process in mind, the answer was that “in our
mind actually it was very structured, from A to Z…” (ibid.). The actual process
was described as “like a spaghetti and meatballs sauce thing because it was very
difficult to divide the few things, the techniques from the aesthetics, from the
problems the users have...” (ibid.).
Finding users was the initial difficulty for the students: “I would say to find
users took a very long time......quite some weeks actually. I would say the first one
and a half months went on research...” (ibid.). The project duration was 12 weeks.
Seven users were identified but they were not able to contribute as expected. The
designers tried to use cultural probes but the resultant material was sparse and
unsatisfactory. Workshops and brainstorming sessions were also planned but found
unworkable. The actual strategy adopted was to video record interviews. This
proved to be time consuming and created ethical and welfare problems: “[The
patients] get very tired and just speaking for one hour is a project that takes more
or less one day for them [...] For them to go from the kitchen to a table two metres
away would take a lot of energy. And the interviews were very calm, relaxed [but]
it was very difficult for them. You could see the effort they were making. We did
not want to take a lot out of them, we did not want to kill them in our
interviews....” (ibid.).
To work around these problems the designers interviewed medical personnel
and they used themselves as subjects in testing parts of the proposed designs. The
problem of finding an optimum location for the nasal catheter (a thin tube
connected at one end to the portable unit) was one which side-stepped difficulties
of working with their users directly. This was a matter of prototyping the range of
positions which eliminated obstruction caused by the tube connecting from the
nose to the transportable device. This necessitated a vertically mounted tube in
contrast to the side-mounted tubes of standard designs. Inclusive design and
standard ergonomic practice warn against using the designer as substitute for the
user. However, in the case of frail and weak users this goal was difficult practically
and ethically. The designers used a breath restriction device - essentially breathing
through a narrow tube - to simulate the reduction in oxygen intake. This allowed
them to understand to some degree their users´ difficulties. A physiotherapist
provided insight concerning the practical ergonomics of the design. This focused
on minimising interference with thoracic muscle movement due to the weight of
the portable unit. Having settled on a design offering the best compromises, the
design was shown to what the designers called their “super-user”. This person was
the most committed individual in the user-group who, coincidentally, had a better
170 Herriott
than average understanding of the design process. They were able to spend more
time and to more fully explain their needs and experiences. In this case the super-
user was atypical because of their engagement rather than because they were a
more extreme case of COPD. After this consultation a full-size mock-up was
prepared and the project thus completed.
17.3.2 The Four Municipalities Project

Four municipalities (Aarhus, Esbjerg, Næestved and Roskilde) in Denmark formed
a consortium to develop devices to increase the independence of elderly people
with a variety of medical conditions. The intention was that the concepts should
also not require additional manpower or, at best, reduce the need for it. The themes
of the project were 1) the toilet, 2) outdoor and indoor mobility, 3) home training
and personal care and 4) mobility for obese citizens. The project was managed by
CAT in Roskilde, Denmark. CAT, formerly known as the Centre for Applied
Technology, is a privately owned company which helps entrepreneurs and
innovators to find the capital and competences required for converting ideas into
enterprises. The design was carried out by CPH Design from Copenhagen (the
initials “CPH” refer only to the official code for Kastrup Airport).
The design process consisted of the following steps: a start-up process,
research, analysis, conceptual development and evaluation. The start-up process
involved assembling expertise from the Copenhagen Business School,
anthropologists from Aarhus University, the Danish Centre for Assistive
Technology, the Danish Technological Institute and interest groups representing
the elderly and handicapped. The stakeholders thus had an extensive role in a
product’s development as per ID’s tenets.
Figure 17.2. Body drier concept drawing

The research phase was intended to involve a process approximating to a form

of co-design where users themselves wielded pens and cameras to document their
activities. This turned out not to be possible. The project manager Bente Rugaard
Thorsen described the problems in this way: “I thought I could use a lot more
creative methods. I wanted to bring the elderly in for interviews where I could ask
them to draw their daily life, to ask them to make pictures of what they saw as
being difficult for them...some of them could not hold a pen” (Rugaard Thorsen,
2011). Another device, the user diary, also had to be dropped for the same reason.
An extended explanation of the problem was provided: “Having studied learning
and innovative processes in the bank sector, I used drawing and sketching, but
these people [in the study process] were not able to write their own name; many of
them suffer from spasms or suffered from paralysis which often affected both sides
of their bodies....or had lots of psychological problems…” (ibid.). Verbal
communication was sometimes hindered by the effects of stroke where “you have
difficulty structuring your speech... or they could have difficulties remembering
and explaining…” (ibid.). The participants’ physical weakness limited the time for
discussions and taking part in processes such as workshops which were viewed as
disruptions to daily routines. Other methods had to be used to “get a glimpse of
the challenges of their daily life…” (ibid.).
An anthropological approach was taken in response to these problems. This
required interviewing and observing approximately 50 participants. The very
particular problem of how to convey very personal, intimate problems related to
hygiene was encountered. “It would have been too intimate for these people to sit
in a workshop, talking about their own problems and to be analysed openly in front
of care-givers and also the companies [who would produce the products]” (ibid.).
Thus, the strategy of using personas was adopted. While this is less than ideal it
allowed for the anthropologist to communicate the needs of the users to the
designers. Interview data was generalised and combined into figures that
represented typical problems of each sub-group. The information was presented
verbally and through posters. Idea cards were used to supplement this information.
The investigation phase generated 140 ideas. From these the designers selected
12 core concepts to be refined: a modified shower-head, a virtual rehabilitation
training programme, a Scandinavian-style treadmill with a virtual street view, a
powered zipper, a harness for rising and sitting, a body dryer, a heated bathrobe,
mark-up stickers for training with wheelchairs, a uridom and a folding bath. Two
of the concepts - not listed - have been taken on for further development and for
reasons of commercial sensitivity have not been presented publicly. There was
blending of the research, analysis and conceptual development stages. To quote the
project manager: “It’s never a linear process, but in a broad sense, we followed our
planned design process” (ibid.).
This project showed how both psychological inhibitions and physical frailty
were barriers to gaining clear user insights. Specifically, when dealing with frail
individuals there is an extra complication when the physical problems involved
relate to matters of bodily privacy. The process highlighted a paradox. Subjects
reported that it was humiliating to ask for help but in order to reduce their need for
help, they must first agree to be helped. And not only helped in the sense of
momentary assistance, but helped in the form of discussing their problems.
172 Herriott
17.4 Discussion
It has been taken as given that when designers wish to include users then those
users will be willing and able to be included. These cases show that the ideal of
maximising user input into ID processes is met by the difficulties those users
themselves have. Both the Four Municipalities project and Breathe Easy show that
the conditions which make it necessary to think inclusively can make inclusivity
difficult to implement. The AT user group is not merely slightly or moderately less
physically capable than the “standard user group.” They are sometimes much less
able to be active contributors to the design process aimed at helping them and
which depend on their input. This does not mean that inclusive design is not
possible. Rather a change in approaches is needed, what could be termed an
economical rationing of user input and an emphasis on less intrusive methods. The
Breathe Easy project used physiotherapists as surrogates: they had the means to
communicate observations that a designer would not have time to make. But they
are not the users we think of when thinking user-centred design. These projects
also demand that we re-evaluate certain assumptions. For example, Porter and
Porter (2002) list eight fallacies regarding ergonomics. The first fallacy (and the
only one relevant here) is: “the design is satisfactory for me - it will therefore be
satisfactory for everyone else.” This is generally a good assumption but it is not an
absolute. The Breathe Easy team were faced with either self-testing their designs or
not testing at all. Their cases indicate that there are some instances where designers
can use their experience to make decisions which are (perhaps) less than ideal but
which can still produce work which, by any objective standard, is an appreciable
improvement on the alternative. To be abstracted from this is that we may need to
counterbalance inclusive design’s centripetal force that drags users into the
process. Designers in the end have to make design decisions. They may have to use
their understanding and judgement, distilled from their other work, to provide
substitutes and surrogates for the absent, unwilling or disabled user.
In the case of the Four Municipalities project, the intention to use a co-design
process was unfeasible. Again, designers had to take on work that ideally users
would participate in. The problems involved recall those described by Boyd-
Graber et al. (2006) where a “modified participatory design approach was used in
which proxies, that is, speech-language pathologists who work with aphasic
individuals, assumed the role normally filled by users.” Underwriting the validity
of their work was the use of extensive interviewing and the use of anthropological
professionals to guide this. This extensive interviewing provided a rich seam of
information with which to cross-reference the findings. Such work is expensive
and time consuming and it was made possible only by a favourable context for
inclusive design (patient and well-funded clients). Other designers might not be so
fortunate.
ID and AT have much in common but are not the same. This paper has shown
how common linkage of the two elides the differences in methods that are
appropriate to designers in the respective fields. Those who are too ill to do the
things we take for granted in the “broader mainstream” are by definition a
distinctly different group. In metaphorical terms, the toolbox for ID and AT is the
same. However, AT requires certain tools to be used in different ways. Designers

must adjust their expectations of what can be achieved with users according to the
nature of the capability loss in question.
Regarding future work, it might be fruitful to consider re-engineering the
inclusive design model to account for the differing capabilities of the target user
group. In this case we might want to find out how to focus user-involvement more
on the key functionalities and less on areas where surrogates can be found to
provide information. One temptation in response to an inclusive design problem is
that of devising more guidelines. Such guidelines as designers need are better
created on a case by case basis. There is tremendous diversity in capability loss and
a general guideline would, in all likelihood, not be able to accommodate this. If
anything, this work shows the difficulty inherent in a priori formulations
concerning problem-solving.
17.5 References
Barbiani M (2011) Interview with author, conducted April 4, 2011
Bichard J-A, Langdon PM, Coleman R (2007) Does my stigma look big in this? Considering
acceptability and desirability in the inclusive design of technology products. In:
Stephanidis C (ed.) Universal access in human-computer interaction: Coping with
diversity, 4th International Conference on Universal Access in Human-Computer
Interaction (UAHCI 2007), Springer, Berlin, Germany
Boyd-Graber JL, Niklova SS, Moffat KA, Kin KC, Lee JY, Mackey LW et al. (2006)
Participatory design with proxies: Developing a desktop PDA system to support people
with aphasia. In: Proceedings of CHI 96 Workshop on Human Factors in Computer
Systems, SIGCHI, New York, US
BSI (2005) Design management systems - part 6: Managing inclusive design - guide. British
Standards Institution, London, UK
Center for Universal Design (1997) The principles of universal design. North Carolina State
University, Raleigh, NC, US
Cook A, Hussey SM (2002) Assistive technologies - principles and practices. Mosby Inc, St
Louis, MO, US
Coleman R, Clarkson PJ, Dong H, Cassim J (2007) Design for inclusivity: A practical guide
to accessible, innovative and user-centred design. Gower, Aldershot, UK
Clarkson PJ, Coleman R, Hosking I, Waller SD (2007) Inclusive design toolkit. Engineering
Dhiensa J, Machin C, Stone R (2005) Assistive technology: Going beyond the disability. In:
Proceedings of the International Conference on Inclusive Design and Communications
(INCLUDE 2005), London, UK
Green WS, Jordan PW (2002) Pleasure with products. Taylor & Francis, London, UK
Jeffery PK (1998) Structural and inflammatory changes in COPD: A comparison with
asthma. Thorax, 53(2): 129-136
Linnott A (2011) Influencing the assistive technology marketplace. In: Proceedings of the
London, UK
Mayagoitia RE, Kitchen S, Harding R, King A, Turner-Smith A (2006) User-centred
approach to the design and evaluation of a stair climbing aid. In: Clarkson PJ, Langdon
PM, Robinson P (eds.) Designing accessible technology. Springer, London, UK
174 Herriott
Mountain GA, Ware PM, Hammerton J, Mawson SJ, Zheng H, Davies R et al. (2006) The
Smart project: A user-led approach to developing applications for domiciliary stroke
rehabilitation. In: Clarkson PJ, Langdon PM, Robinson P (eds.) Designing accessible
technology. Springer, London, UK
Orpwood R, Gibbs C, Adlam T, Faulkner R, Meegahawatte D (2004) The Gloucester smart
house for people with dementia - user interface aspects. In: Keates S, Clarkson PJ,
Langdon P, Robinson P (eds.) Designing a more inclusive world. Springer, London, UK
Orpwood R, Chadd J, Howcroft D, Sixsmith A, Torrington J, Gibson G et al. (2008) User-
led design of technology to improve quality of life for people with dementia. In: Langdon
PM, Clarkson PJ, Robinson P (eds.) Designing Inclusive Futures. Springer, London, UK
Pullin G (2009) When design meets disability. MIT Press, Cambridge, MA, US
Porter M, Porter CS (2002) Occupant accommodation: An ergonomics approach. In:
Happian-Smith J (ed.) An introduction to modern vehicle design. Butterworth
Heinemann, Oxford, UK
Reed DJ, Monk A (2006) Design for inclusion. In: Clarkson PJ, Langdon PM, Robinson P
(eds.) Designing accessible technology. Springer, London, UK
Rugaard Thorsen B (2011) Interview with author, conducted May 31, 2011
Story M, Mueller JL, Mace R (1998) The universal design file: Designing for people of all
ages and abilities. Center for Universal Design, North Carolina State University, Raleigh,
NC, US
Torrens G (2010) Assistive technology design: A mixed methods approach. Loughborough
Design School, Loughborough University, Loughborough, UK
Chapter 18
What is Good Design in the Eyes of Older

Users?
N. Goddard and C. Nicolle
18.1 Introduction
With the population of older consumers increasing and with the recent changes in
legislation and attitudes towards this group, there have been corresponding changes
in product design practice and a growing attempt to adopt an inclusive design
approach. This recognises that people can become excluded from using products,
services or environments if the needs and capabilities of all potential users are not
taken into account. The inclusive design approach has developed from
collaborations between industry, designers and researchers. One major influence
in this area is the i~design project, whose definition is simply that “inclusive
design is better design” (EDC, 2011). The Inclusive Design Toolkit website, a key
output from the i~design project, states that a successful product must be
“functional, usable, desirable and ultimately profitable” and that a key to good
design is to reduce the demand on the user when capabilities decline with age or
disability (EDC, 2011).
It is also important to consider more emotional aspects, such as social
acceptability and whether the potential user would actually want to use or be seen
using the product (Keates and Clarkson, 2003). Other authors also emphasise that
whilst inclusive design research and practice to date have focused primarily on the
physical accessibility and usability of products, a better understanding is required
of people’s emotional needs, such as social acceptability and desirability of
products (Coleman et al., 2007; Lee, 2010). Similar views regarding the required
shift in design focus are reflected in a number of other sources: the need to
consider the less tangible human factors such as identity, emotion, delight and self-
expression (Cassim et al., 2007); simplicity, aesthetics, pleasure, personality,
conspicuousness and fashion (Pullin, 2009); the product’s visual appearance (Crilly
et al., 2004); creating pleasurable experiences (Jordan, 2000; Demirbilek and
Sener, 2003); and the importance of the emotional aspects of design for a
successful product (Norman, 2004), as well as needs related to specific cognitive
conditions (e.g. Baumers and Heylighen, 2010).

176 Goddard and Nicolle
However, knowledge about what is good design can only arise by cooperation
between designers and the end users; designers themselves cannot always judge
what is good without understanding the point of view of the users (Heylighen and
Bianchin, 2010). Therefore, the objective of the current study was to take a user-
centred approach to investigate what makes good product design in the eyes of
older people themselves and how their criteria for good design compare with those
considered by designers and researchers.
18.2 Methods
The first stage involved an ethnographic approach, combining semi-structured
interviews and observation in people’s homes, enabling an understanding of older
users’ individual opinions and use of everyday products in their natural
environment. All procedures were approved by the University’s Ethical Advisory
Committee.
Participants were either emailed or handed (as per their preference) a pre-
interview questionnaire to complete and return prior to the start of the interview.
The first part of the questionnaire collected personal background information to
provide context for the interview and analysis. The second part explained a pre-
interview task: to think about one product they consider of good design and one
product they consider of bad design. Whilst a detailed interview schedule was
developed to enable a certain level of consistency of questioning across interviews,
the interviews were semi-structured in order to provide flexibility to follow the
lead of the participant and their own terminology, and also to follow up on any
unexpected line of thinking.
To enable some degree of quantification of responses, a shuffle card exercise
was introduced which required each participant to prioritise a set of 30 criteria,
written on cards, by dividing them into three piles in terms of importance to them
for good product design. An example of a ‘most important’ pile is shown below in
Figure 18.1.
Figure 18.1. Shuffle card exercise to prioritise criteria

What is Good Design in the Eyes of Older Users 177
During the interviews, participants were observed using the products and any
difficulties encountered with them. The benefits of taking a more ethnographic
approach were proven to be as follows: participants were able to demonstrate how
they used products in their normal environment and the difficulties they
experienced; participants’ non-verbal cues could be observed to add understanding
and context, e.g. laughing or tone of voice; and observation revealed gaps between
what people said and what happened in reality.
Transcripts of all interviews were made in order to provide a complete and
accurate record of the data collected and to enable structured analysis. An initial
review of the transcripts was carried out to input to the design of the online
questionnaire, hosted on SurveyMonkeyTM. This formed the second stage of the
study and enabled a certain degree of validation of the stage one findings. The
language used and the themes covered in the online questionnaire were based on
both the interviewees’ comments and themes from the literature review.
The data from the interviews and the online questionnaire were coded and
analysed according to key themes, their source and frequency of occurrence, using
the qualitative data analysis software NVivo.
18.3 Results
Thirteen interview sessions were conducted amongst people aged 65 or over,
lasting between 45 and 90 minutes, including a mix of one-to-one interviews,
paired interviews with couples and a mini group of three friends. This resulted in a
total sample of 22 participants. During the interviews, participants were observed
demonstrating the products and any difficulties encountered with them.
Details of the age, gender, living arrangements and impairments of the
interview participants are summarised in Table 18.1.
Table 18.1. Interview participant details

Total Male Female
Age
65-74 7 1 6
75-84 13 7 6
85+ 2 1 1
Living arrangements
Live alone 7 2 5
Live with partner 15 7 8
Impairments
(moderate or severe)
Vision 10 4 6
Dexterity (arthritis) 5 2 3
Mobility 4 2 2
Multiple impairments 8 4 4
For the online questionnaire, a sample of older users was achieved by obtaining
permission to post a link on a number of University of the Third Age (U3A)
websites across the East Midlands. 72 respondents completed the questionnaire,
although only 54 answered the demographic questions at the end of the
questionnaire (Table 18.2).
Table 18.2. Online respondents (n = 54 out of 72)

Total 54
Male 14
Female 40
Age
55-64 19
65-74 25
75-84 9
85+ 1
The key themes that emerged from analysis of the interview data and the
responses to the online questionnaire were based on the frequency of responses and
are summarised in Figure 18.2. High level themes, which were then broken down
into finer categories, included the elements of good and bad design; elements in
purchase decision; comments on aesthetics vs. function; role of family, friends and
other people; tricks, solutions and adaptations; and reactions to Good Grips
products.
Figure 18.2. Key themes from the research

The most important criteria for good design amongst older users in this study
are that products are easy to use, they do the job they are supposed to well and are
simple to understand. The most likely problems that older users experience with
products, and which are therefore seen as elements of bad design, are difficulties
getting into some types of packaging and the heavy weight of some products.
These findings are not surprising; however, the section below discusses some
interesting results which shed a new light on these issues.
18.4 Discussion
18.4.1 Aesthetics Versus Usability
From detailed analysis of data from both the interviews and open-ended online
questions, it would appear that many older people would prefer, ideally, to have
products that were not just easy to use but also looked good - but importantly, only
if usability has been delivered. This idea links to Maslow’s hierarchy of needs
(Maslow, 1987) which is often used to help understand consumers’ requirements
from products. According to this model, once the needs at the lower levels, e.g.
safety and comfort, have been satisfied, emphasis can shift to needs at the higher
levels, that is, towards the more emotional attributes of a product. Jordan adapted
Maslow’s model to a Human Factors perspective, creating a new three-level model
with ‘functionality’ at the lowest level, ‘usability’ in the middle and ‘pleasure’ at
the highest level (Jordan, 2000). Using this model to interpret the current results,
users’ basic needs of functionality (performs the tasks for which it is intended) and
usability (easy to use) do not appear to be met. Many are experiencing difficulties
and are therefore not able to progress to the higher levels to meet their emotional
needs. If the more functional needs were being addressed successfully then the
‘pleasure’ needs, including looks, would become increasingly important.
The literature review revealed that several authors believe that the focus for
design should shift to delivering fulfilment at the higher levels, for both users in
general and older users (e.g., Crilly et al., 2004; Lee, 2010). However, the results
from this study would indicate that product design may be leaving older users
behind by not delivering successfully at the functional level for these groups.
These results support the view that some designers may have become overly
concerned with the aesthetics of product interfaces, resulting in problems being
caused for people with impairments (Noonan, 2007).
18.4.2 Tricks and Solutions for Packaging Problems

The current study indicates that difficulties encountered in getting into different
forms of packaging are not uncommon, and for most (but interestingly not all)
people a source of irritation. In particular, it would appear that many people have
problems using can and jar openers and with opening child-safe bottle tops.
Difficulties and alternative solutions identified in the current study are very similar
to those revealed in the packaging research (e.g. Yoxall et al., 2010), that is, using
either physical strategies, for example using a tool such as a knife, or social
strategies, for example asking a relative (see next section).
In some cases participants appeared to be proud of these solutions and were
quite animated whilst explaining them. Is it possible that such problem solving
could be beneficial in keeping the older mind active and in giving the user a sense
of pride? The packaging study mentioned above also refers to the pride and
achievement felt by users regarding their ability to solve problems (Yoxall et al.,
2010). This is possibly an interesting area for further research, although the
negative consequences of failing to solve a poor packaging ‘puzzle’ would almost
certainly be greater than the gains of solving that puzzle.
18.4.3 Involvement of Other People

It appeared from the current study that a significant influence on the products used
by older consumers comes from other people rather than the users alone. Products
were often bought for them by others, or were seen being used by others, or were
recommended by others. Other studies have found that people in the 76+ age
group were more likely than the younger 65-75 age group to rely on others to
choose or purchase new products for them (Burrows et al., 2010).
For one participant in the current study, when her son gave her a can opener for
Christmas he demonstrated to her how easy it was to use, but when she later tried
to use it herself she could not use it. Therefore, an important influence on what
other people buy for older users would be the perception the purchaser has of what
is usable by, or will be liked by, the user - that is, what they believe makes good
design for older users if they are not older users themselves. Further research
amongst these purchasers would enable a greater understanding.
This study indicated that there is a family (or friend) social dynamic involved
that may play an important role in the lives of older people, particularly those
living alone, for example when a son or daughter is called upon to help out with
problems experienced. Burrows et al. (2011) also found that, given the choice,
older adults often decide to involve other people in the various stages of their
interaction with new technology. There may be a strong need by older people for
the social interaction that these ‘difficulties’ bring with them, perceived or
otherwise, and the authors believe, in line with Burrows et al. (2011) that more
understanding is required about this context. Further research might therefore be
valuable to get a better understanding of this role, its implications and the extent to
which product design problems and affordances ‘encourage’ interactions between
family members or neighbours.
18.4.4 The Effect of Familiarity on Purchase

Several participants were using the same products over many years and because
they were so used to a product they would plan to buy the same one again. It is
possible that it is more typical for the older generation than subsequent more
‘disposable’ generations to hold on to products for as long as possible. As a result
of looking for this familiarity with what they already know, it is possible that older
people may avoid innovative designs. This poses a challenge for designers: what
can they do to ease the transition for these older users from their very familiar
products to the unfamiliar, newer product interfaces? Previous experience with
similar products is a strong predictor of usability, and those products that help the
user make a reference to the same function on another more familiar device should
perform better than those that do not (Lewis et al., 2006; Langdon et al., 2007).
Furthermore, what innovators see as providing personal benefit may not be seen as
such by the older person. The same can be said by the family of an older person -
if a family member purchases a new product to replace the old, worn out one, the
older person might have preferred to ‘battle on’, rather than change the way they
do things. How do we encourage users to ‘battle on’ with the new product long
enough to recognise the benefits?
18.4.5 Expectations
There was some indication from the interviews that, amongst this older generation,
expectations for products to work perfectly are low. In addition, a few participants
were quite accepting of the fact that in some instances they would not be able to
understand how to use the products: “Why should we get our knickers in a twist
because we can’t understand everything that’s modern?” Having to find
alternatives or make adaptations in such situations does not appear to be
troublesome for many participants. One possible explanation is that the older
generation are used to having to ‘make do’ and to adapt existing products to make
them usable and to last.
However, the Baby Boomer generation are just about to enter the 65+ ‘older
user’ category. They are considered to have two distinctive characteristics,
individualism and liberalism, which are likely to affect their attitudes to products
and product design (Huber and Skidmore, 2003). Compared with the previous
generation, that is the current ‘older users’, they may be more likely to complain
about products and to expect their individual wants and desires to be satisfied.
Whilst the older participants in the interviews appeared to be more accepting of
design problems, the younger-old are likely to be less tolerant, more demanding
and therefore more likely to complain when they have problems.

Despite the volume of research conducted on product usability and accessibility,
results from the current study into what makes good design for older users would
indicate that the basic functional needs are in fact still not being met for this user
group. In particular, older people are experiencing difficulties with ease of use,
packaging and weight of products. As explained by Maslow’s hierarchy of needs,
the lower level, basic needs have to be addressed before users’ higher level
emotional needs can be met. Therefore, the belief that appears to exist from the
literature review, that the focus for product design needs to move away from the
basics to the more emotional aspects of product design, would not yet appear to be
advisable when considering older users. The recommendation is for designers to
continue to focus on the basics of accessibility and usability of mainstream
products for older users, but designs should of course aim to excel in both function
and desirability.
A number of other key themes emerged from this study as possibly warranting
further investigation. There was some indication that the solutions the users had to
find in overcoming difficulties getting into product packaging were giving them a
sense of pride. Further research might be useful in gaining a greater understanding
of the value of this problem-solving for older users, the possible benefits for
keeping the older mind active and the implications for product design.
In many cases, other people such as friends or relatives are purchasing
everyday products for older users. It might therefore be useful to get a greater
understanding of what the purchasers think is good design for the users for whom
they are buying, particularly when they may be significantly younger than the users
themselves. Another aspect relating to other people is the social dynamic that
exists when others are purchasing products for older people, or are called upon to
assist them with the products they use. Further research might be useful to
understand the role such situations play in instigating social interactions and the
importance these have in older people’s lives.
Another key theme identified is the desire by many older users to buy products
with which they are already familiar. Further research might help understand the
extent to which this might hinder their acceptance of innovation and what
designers can do to minimise this conflict.
However, the most important message from the current study is that the
fundamental need to get the basics right for older users will be increasingly critical
as the current Baby Boomer generation are beginning to enter the ‘over 65’
category. Compared to the current group of older users they are likely to be more
demanding, less tolerant and more prone to complain about any shortfall in product
design, whether in functionality or style, in meeting their needs and aspirations.
18.6 References
Baumers S, Heylighen A (2010) Harnessing different dimensions of space: The built
environment in auti-biographies. In: Langdon PM, Clarkson PJ, Robinson P (eds.)
Designing inclusive interactions. Springer, London, UK
Burrows A, Mitchell V, Nicolle CA (2010) Out-of-box experiences: an opportunity for
inclusive design. In: Proceedings of 5th Cambridge Workshop on Universal Access and
Assistive Technology, Cambridge, UK
Burrows A, Mitchell V, Nicolle C (2011) Designing in social benefits. In: Proceedings of
the International Conference on Inclusive Design and Communications (INCLUDE
2011), London, UK
Cassim J, Coleman R, Clarkson PJ, Dong H (2007) Why inclusive design? In: Coleman R,
Clarkson PJ, Dong H, Cassim J (eds.) Design for Inclusivity. Gower Publishing,
Aldershot, UK
Coleman R, Topalian A, Clarkson PJ, Dong H (2007) The business case. In: Coleman R,
Clarkson J, Dong H, Cassim J (eds.) Design for inclusivity. Gower Publishing, Aldershot,
UK
Crilly N, Moultrie J, Clarkson PJ (2004) Seeing things: Consumer response to the visual
domain in product design. Design Studies, 25(6): 547-577
Demirbilek O, Sener B (2003) Product design, semantics and emotional response.
Ergonomics, 46(13/14): 1346-1360
EDC (2011) Inclusive design toolkit. Cambridge Engineering Design Centre, University of
Cambridge, Cambridge, UK. Available at: www.inclusivedesigntoolkit.com (Accessed
July 2011)
Heylighen A, Bianchin M (2010) Can crap design be inclusive? The case for deliberative
design. In: Proceedings of 5th Cambridge Workshop on Universal Access and Assistive
Technology, Cambridge, UK
Huber J, Skidmore P (2003) The new old. Why baby boomers won’t be pensioned off.
DEMOS, London, UK
Jordan PW (2000) Designing pleasurable products. Taylor & Francis, London, UK
Keates S, Clarkson PJ (2003) Design exclusion. In: Clarkson PJ, Coleman R, Keates S,
Lebbon C (eds.) Inclusive design: Design for the whole population. Springer-Verlag,
London, UK
Langdon P, Lewis T, Clarkson J (2007) The effects of prior experience on the use of
consumer products. Universal Access in the Information Society, 6(2): 179-191
Lee Y (2010) Development of the social implications of inclusive design and some thoughts
on the next steps. In: Proceedings of the 5th Cambridge Workshop on Universal Access
and Assistive Technology, Cambridge, UK
Lewis T, Langdon PM, Clarkson, PJ (2006) Investigating the role of experience in the use of
consumer products. In: Clarkson PJ, Langdon PM, Robinson P (eds.) Designing
accessible technology. Springer, London, UK
Maslow AH (1987) Motivation and personality, 3rd edn. Harper and Row, NY, US
Noonan T (2007) The overlooked consumers: A discussion paper examining the access,
challenges and emerging possibilities for consumer electronics and home appliances.
Australian Human Rights Commission, Sydney, Australia
Norman DA (2004) Emotional design: Why we love (or hate) everyday things. Basic Books,
NY, US
Yoxall A, Langley J, Musselwhite EM, Rodriguez-Falcon EM, Rowson J (2010) Husband,
daughter, son and postman, hot-water, knife and towel: Assistive strategies for jar
opening. In: Langdon PM, Clarkson PJ, Robinson P (eds.) Designing inclusive
interactions. Springer, London, UK
Chapter 19
Equal Access to Information? Evaluating

the Accessibility of Public Library Web
Sites in the State of Maryland
J. Lazar, B. Wentz, C. Akeley, M. Almuhim, S.
Barmoy, P. Beavan, C. Beck, A. Blair, A. Bortz, B.
Bradley, M. Carter, D. Crouch, G. Dehmer, M.
Gorman, C. Gregory, E. Lanier, A. McIntee, R.
Nelson Jr., D. Ritgert, R. Rogers Jr., S. Rosenwald,
S. Sullivan, J. Wells, C. Willis, K. Wingo-Jones and
T. Yatto
19.1 Introduction
Public libraries in the United States have a long and proud tradition of providing
access to information for all residents. Public libraries have been the equaliser -
providing access to books, and both printed and electronic information, regardless
of race, gender, religion, economic status, or disability. Since the mid-1990s,
public libraries have also been providing direct access to the Internet for patrons
who come to visit. And since that first burst of access to the Internet within public
library buildings, libraries have been pushing to provide access to their resources
through the Internet, so that patrons can search library catalogues, reserve
resources and renew materials, and even access digital libraries of documents, all
wherever the patrons happen to have Internet access - at home, work, or using
mobile phones (Jaeger et al., 2011).
People with disabilities often use library resources, but since transportation to
public library buildings is often a challenge, many people with disabilities may
prefer to use the resources of the library in electronic format. Therefore, it is
important to examine the accessibility of public library web sites and determine
whether or not the web sites are in compliance with public laws related to web
accessibility. This chapter presents a research study where the home pages of all 24

186 Lazar et al.
public library system web sites in the state of Maryland were evaluated for
accessibility.
19.2 Related Literature

19.2.1 Libraries and Accessibility
Public libraries in the United States actually have a long history of involvement in
the area of accessibility. In the past, this was often focused on physical issues, such
as wheelchair access to library buildings, and alternative formats for printed
materials (Lazar et al., 2011). For instance, the Library of Congress’s National
Library Service (NLS) for the Blind and Physically Handicapped, founded in 1931,
focuses specifically on development of and distribution of library resources in
alternative formats, such as large-print, braille, and audio (Dziedzic, 1983). The
American Library Association notes that equal access to information is extremely
important and a significant challenge facing modern libraries (American Library
Association, 2011). So there is a history of concern about both physical building
accessibility, as well as the accessibility of library content material. However, there
is less of a documented tradition about web accessibility for public library
resources. This concern needs to include both the public library web sites, and any
digital libraries offered through the library web site to library patrons (Bertot et al.,
2006). These digital libraries and other services, since they are typically provided
by outside firms, are sometimes the hardest to manage in terms of accessibility,
since they are out of the control of the public library (Byerley and Chambers,
2002). Recent research documents that many of these digital libraries continue to
be inaccessible (Tatomir and Durrance, 2010). Typically the only way to enforce
web site accessibility for externally-managed digital libraries is through closely-
monitored procurement/acquisition processes.
There is surprisingly little data about the accessibility of public library web sites in
the United States. The one previously-published study examining the accessibility of
public library web sites, during the year 2000, found that only 14/74 of public library
home pages were fully accessible (Lilly and VanFleet, 2000), and that data is more than
10 years old. There have been more published studies about the accessibility of
university library and school library web sites, such as Spindler (2002) and Comeaux
and Schmetzke (2007), who both found that a majority of the home pages that they
evaluated were not accessible. The largest annual data collection effort related to US
public libraries and the Internet since 1994, known as the Public Libraries and the
Internet (and more recently known as the Public Library Funding and Technology
Access Survey), does not include any questions related to technology accessibility
(http://www.plinternetsurvey.org/). Therefore, the authors decided that it was important
to examine public library web site accessibility within the state of Maryland.
Equal Access to Information? 187
19.2.2 Library Structure in Maryland

In the state of Maryland, the Division of Library Development and Services is a
part of the Maryland State Department of Education, serves a central coordination
role, and administers state and federal programs related to libraries. However each
of the 24 public library systems is administered by the county in which it is located
(Maryland State Archives, 2011). Note that Baltimore City actually is not part of a
county, so it is considered a separate administrative entity, and has a separate
library system (the Enoch Pratt Free Library) from the Baltimore County Public
Library. Typically, each county public library system has multiple branches, but
one web site. There are actually multiple laws in place that require public library
web sites to be accessible. For instance, the Maryland Information Technology
Nonvisual Access (MD IT NVA) Regulatory Standards require all state
government web sites to be accessible (Maryland State Department of Information
Technology, 2011), and Title II of the Americans with Disabilities Act (a federal
law) covers state and local government services. The technical specifications for
web sites, from the Maryland Information Technology Nonvisual Access
regulations, are the same as the Federal Section 508 regulations, which cover
Federal information technology.
19.2.3 Website Accessibility

Users with perceptual and motor impairments often use alternate input or output
approaches to access web sites. For instance, low-vision users may utilise screen
magnification. Blind users typically access web sites through the use of text-to-
speech (or screen reader) software, which reads the content of a web page in an
audible manner to the user in a linear fashion. Examples of screen reader software
include JAWS, WindowEyes, NVDA, and VoiceOver. Screen reader software is
the dominant method of access because Braille literacy is extremely low among
blind people (National Federation of the Blind, 2011) . Users with limited dexterity
may use alternative keyboards, no pointing device, or speech recognition. Users
who have spinal cord injuries and/or paralysis may utilise speech recognition or
eye tracking. Deaf or hard of hearing users may rely on captioning for videos, or
transcripts for audio. Web developers, as well as public policies, tend to focus on
the accessibility needs of people with perceptual and/or motor impairment, rather
than cognitive impairment, because, while there is nearly a 30-year history of
interface design for people with perceptual and/or motor impairment, there is a
much smaller history of interface design for people with cognitive impairment, and
there are currently no design standards for such people (Lazar, 2007). The
international design standards for web content are the Web Content Accessibility
Guidelines (WCAG), from the World Wide Web Consortium, and the current
version is WCAG 2.0. Web Accessibility regulations in countries such as the US,
Canada, UK, and Australia, are all based on the WCAG. Problems arise when web
site interfaces are not designed to standards (such as Section 508 in the US or the
188 Lazar et al.
Equality Act in the UK). Poor design results in web site content being inaccessible
to users with disabilities, creating exclusion and discrimination.
19.3 Research Methodology

Expert inspections of web site interfaces were the methodology chosen for this
research. Automated software (such as A-Prompt, Deque WorldSpace, and
WebXACT) is often used for accessibility testing, but automated evaluations are
not as accurate as expert inspections or user-based testing. This is because some of
the most common accessibility problems (such as link titles or alternative text that
do not fit the context) cannot be determined in an automated fashion. The most
effective form of accessibility evaluation is done with a screen reader, where
experts without visual impairments evaluate compliance with technical and legal
standards for web sites (Mankoff et al., 2005). The accuracy of this method
increases when multiple evaluators inspect the same web site interface (Lazar et
al., 2010).
Expert inspections tend to focus on technical accessibility, rather than true
usability. On the other hand, usability testing, involving people with disabilities,
provides a higher understanding of ease of use. Usability testing is task-focused,
where users attempt to complete a series of tasks. However it is not inspection
focused. As a result, usability testing may not ascertain compliance with the law,
since the users may not actually be able to test all of the various components of the
regulations as a part of their tasks. Due to the various strengths and weaknesses,
the combination of expert inspections and usability testing with people with
disabilities is ideal, but it is not performed often enough, due to the time and cost
involved.
The expert evaluations that were conducted utilised the web site accessibility
standards of Section 508 (1194.22) of the US Rehabilitation Act, which comprise
16 guidelines, identified as paragraphs “a” through “p” (which are the same
technical standards as from Maryland state law). Expert inspection consisted of an
expert with vision inspecting a web site using a screen reader and a check for
compliance with each specific guideline from Section 508. The process was guided
by the “Absolute Minimum Accessibility Inspection” which was developed by one
of the authors and used effectively in other accessibility evaluations (Lazar et al.,
2010). Each home page was evaluated separately by 5 individuals using JAWS 12.
None of these individuals were blind, but all had previous experience evaluating
web sites for accessibility using screen readers. All evaluators then met, discussed
their individual evaluations, and compiled one meta evaluation, which typically has
a higher level of validity than one only. Each paragraph of the guidelines was
weighted equally. There was no measurement of the number of times that a
specific paragraph was violated, only the existence of a paragraph violation,
because the number of paragraphs violated is considered to be a more accurate
measurement of accessibility (Lazar et al., 2003; Lazar et al., 2010). Table 19.1
presents a list of the guidelines, along with a short description of each (note that the
descriptions are those of the authors, and not a part of the US regulations).
Table 19.1. Description of each of the 16 paragraphs of the Section 508 web guidelines
(a) Text Equivalent (have a text equivalent for any graphical elements)
(b) Synchronised Equivalent Alternatives (have captioned video, transcripts of any

audio, or other alternatives for multimedia)
(c) Use of Colour (colour should not be used as the only method for identifying
elements of the web page or any data)
(d) Organisation (style sheets are encouraged, but users should still be able to utilise a
web page when style sheets are turned off)
(e) Redundant Text Links on Server-Side Image Map and (f) Client-Side Image Maps
(redundant clickable links for server-side image maps, and accessible client-side image
maps are preferred)
(g) and (h) Row and Column Headers (use appropriate headers and mark up to allow
easy navigation of a table)
(i) Frames (title all frames and label all frames for easy identification and navigation,
e.g., use “navigation” “main content” and “search” rather than “top” or “bottom”)
(j) Screen Flicker Frequency (limit or eliminate the use of flickering, which can provoke
seizures)
(k) Text-Only Page Default (if a web page cannot be made accessible, provide an
equivalent text-only page, and make sure it is kept up to date)
(l) Scripting Languages (make sure that equivalents for any non-accessible scripting are
included, e.g., for those who are not using pointing devices)
(m) Linked Plug-In or Applet (if any plug-ins are required, make sure to provide a link
to an accessible version of the plug-in)
(n) Online Electronic Forms (all forms must be properly labelled and accessible)
(o) Method to Skip Repetitive Navigation Links (all web pages should have a link which
allows a user to skip directly to the main content, bypassing any site navigation
information)
(p) Alerts on Timed Responses (if any page responses are timed, the user should be
given the opportunity to indicate that more time is needed)
19.4 Results
19.4.1 Section 508 Paragraph Violations
The evaluations of the home page of each of the 24 county library web sites
revealed that all the home pages violated at least two or more paragraphs of the
Section 508 Guidelines, and some library home pages violated as many as six
paragraphs. Table 19.2 illustrates the category violation for each library home
page.
190 Lazar et al.
Table 19.2. Section 508 paragraph violations of Maryland County Library websites
Paragraph: a b c d e g i j k l m n o p
f h
Alleghany X X X X X
Anne
X X X
Arundel
Balt. City X X
Balt.County X X
Calvert X X X
Caroline X X X
Carroll X X X
Cecil X X X
Charles X X X X
Dorchester X X X
Frederick X X X X X X
Garret X X X
Harford X X X X
Howard X X X
Kent X X X
Montgomer
X X
y
Prince
X X X
George’s
Queen
X X
Anne’s
Somerset X X X X X
St. Mary’s X X X
Talbot X X X
Washington X X X
Wicomico X X X X
Worcester X X X X X X
The most common violations were paragraphs “a,” “n,” and “o.” Paragraph “a”
requires there be a text equivalent for a graphical element on a web page. In other
words, if an image or graphic is used or included on a web page, it should have
alternate text that will be available for screen reader users or any other user who
cannot see the image. Paragraph “n” requires that web site forms and form fields
be accessible and properly labelled. The most common violation of paragraph “n”
occurs when there are not clearly understandable labels associated with form input
fields on a web page. The frustration of inaccessible forms could be illustrated by
imagining filling out an order form with no idea which text box requires the name,
address, phone number, or credit card information. Paragraph “o” requires that
there is a method to skip repetitive navigation links. With web pages often having
many navigational links before the main content is reached, a screen reader that is
reading the content to a user in a linear manner will have to read through each link
before reaching the main content. This can be frustrating and time-consuming. The
common solution is to include a link at the top of every page that will allow users
to skip to the main content of the page. It is important that this link is not hidden
since it can also assist in navigation for users with motor impairments.
19.4.2 Examples of Violations

All of the violations in each category are violations of a particular accessible
design standard. This can best be illustrated through examples of violations that
were discovered on the web sites of various Maryland county library systems. For
example, the Alleghany County public library had three videos on its home page
that had no captions and could not be accessed by the keyboard alone and therefore
had an inaccessible plug-in (violating paragraphs “b” and “m”). Anne Arundel
county’s library system violated paragraph “l” with a drop-down navigation menu
that is not accessible by using a keyboard alone (Figure 19.1).
Figure 19.1. The drop-down menu is not accessible without the use of a mouse
Carroll, Charles, Frederick, and Harford county public libraries all had web
sites that displayed flash slide show content that was inaccessible (with no
alternative). Howard and St. Mary’s county public libraries violated paragraph “n”
with an online form that is improperly labelled. Figure 19.2 shows an accessibility
problem identified on the Howard County Public Library’s home page, where
rotating links are not accessible with a keyboard alone.
192 Lazar et al.
Figure 19.2. The rotating flash events and links are not accessible
Washington County’s web site displayed an online event calendar that was not
accessible, and Wicomico county’s web site used a table that contained no headers
(a violation of paragraphs “g” and “h”). Many of these violations could prevent
individuals with disabilities from accessing all or part of the content that is
available on the library home pages.
19.5 Discussion
The data from this study illustrate the inequity of access to public library home
pages, through an accessibility evaluation of home pages of public library systems
in one state of the United States. It is likely that similar results might be discovered
if the public library web sites were evaluated in other states. With both state and
federal laws requiring web accessibility, and with the American Library
Association highlighting equal access to information as something of significant
importance, this is one area in which a higher level of accessibility is clearly
important and could easily be achieved. The majority of the violations fell within
the categories of alternate text for graphics, labelling of form fields, and the lack of
skip navigation links. These are all very simple changes that could be implemented
with little to no cost on the part of the public libraries. However, the authors
acknowledge that in the current environment, with budgets being cut while there is
a higher demand from patrons for services due to the weak US economy, any new
efforts or initiatives, even simple changes to public library home pages to improve
accessibility, may not be implemented quickly due to the heavy demand on library
staff.
The authors believe that the key to improving web accessibility is awareness
and transparency. A good place to start is by posting a web accessibility statement
on the home page, noting what accessibility features exist on the web site, what
design guidelines were used for ensuring accessibility, what ongoing evaluation
(user testing, expert reviews, or automated reviews) is used to ensure ongoing
accessibility, and who users should contact if they experience any problems related
to accessibility. Research has previously established a relationship between strong
accessibility statements and actual higher levels of web accessibility at the state
level (Rubaii-Barrett and Wise, 2008), although this relationship was not observed
in federal web sites (Olalere and Lazar, 2011). Only three of the county public
library web sites have accessibility statements: Baltimore City, Carroll County, and
Montgomery County. The Baltimore City and Montgomery County accessibility
statements reference applicable laws and policies, but the Carroll County Public
Library web site is extremely vague and merely mentions an effort to make the
web site accessible to everyone. Two other counties, Dorchester and Wicomico,
mention general library accessibility, but there is no mention of web site
accessibility.
There are other ideas for improving public library web site accessibility.
Potentially, web accessibility training could be provided at future meetings or
conferences where many employees of public libraries in Maryland are present. In
addition, we suggest that while expert inspections were used in this study, it would
also be useful to do usability testing involving library patrons with disabilities.
Year-by-year comparisons of web site accessibility can also be helpful to
determine if progress is being made in improving public library web site
accessibility. All of these steps can help raise awareness and hopefully help
improve public library web site accessibility.
19.6 References
American Library Association (2011) Access. Available at: http://www.ala.org/
ala/issuesadvocacy/access/index.cfm (Accessed 9 November 2011)
Bertot J, Snead J, Jaeger P, McClure C (2006) Functionality, usability, and accessibility:
Iterative user-centered evaluation strategies for digital libraries. Performance
Measurement and Metrics, 7(1): 17-28
Byerley S, Chambers M (2002) Web-based library databases for non-visual users. Library
Hi-Tech, 20(2): 169-178
Comeaux, D, Schmetzke A (2007) Web accessibility trends in university libraries. Library
Hi Tech, 25(4): 457-477
194 Lazar et al.
Dziedzic D (1983) Public libraries. In: Cylke F (ed.) That all may read: Library service for
blind and physically handicapped people. US Library of Congress, Washington DC, WA,
US
Jaeger P, Thompson K, Lazar J (2011). Research in practice: The Internet and the evolution
of library research: The Perspective of One Longitudinal Study. The Library Quarterly (in
press)
Lazar J (2007) Introduction to universal usability. In: Lazar J (ed.) Universal usability:
Designing computer interfaces for diverse user populations. John Wiley & Sons,
Chichester, UK
Lazar J, Beavan P, Brown J, Coffey D, Nolf B, Poole R et al. (2010). Investigating the
accessibility of state government web sites in Maryland. In: Langdon PM, Clarkson PJ,
Robinson P (eds.) Designing inclusive interactions. Springer, London, UK
Lazar J, Beere P, Greenidge K, Nagappa Y (2003) Web accessibility in the Mid-Atlantic
United States: A study of 50 web sites. Universal Access in the Information Society, 2(4):
331-341
Lazar J, Jaeger P, Bertot J (2011) Persons with disabilities and physical and virtual public
library settings. In: Bertot J, Jaeger P, McClure C (eds.) Public Libraries and the Internet:
Roles, Perspectives, and Implications. Libraries Unlimited, Santa Barbara, CA, US
Lilly E, VanFleet C (2000) Measuring the accessibility of public library home pages.
References and User Services Quarterly, 40(2): 156-165
Mankoff J, Fait H, Tran T (2005) Is your web page accessible? A comparative study of
methods for assessing web page accessibility for the blind. In: Proceedings of the 23rd
ACM Conference on Human Factors in Computing Systems, Portland, OR, US
Maryland State Archives (2011) Maryland libraries at a glance. Available at: http://www.
msa.md.gov/msa/mdmanual/01glance/html/library.html (Accessed 9 November 2011)
Maryland State Department of Information Technology (2011) Maryland information
technology nonvisual access regulatory standards. Available at: http://doit.
maryland.gov/policies/pages/nva.aspx (Accessed 9 November 2011)
National Federation of the Blind (2011). How many children in America are not taught to
read? Available at: http://www.nfb.org/nfb/braille_initiative.asp (Accessed 9 November
2011)
Olalere A, Lazar J (2011) Accessibility of US federal government home pages: Section 508
compliance and site accessibility statements. Government Information Quarterly, 28(3):
303-309
Rubaii-Barrett N, Wise L (2008) Disability access and e-government: An empirical analysis
of state practices. Journal of Disability Policy Studies, 19(1): 52-64
Spindler T (2002) The accessibility of web pages for mid-sized college and university
libraries. References and User Services Quarterly, 42(2): 149-154
Tatomir J, Durrance J (2010) Overcoming the information gap: Measuring the accessibility
of library databases to adaptive technology users. Library Hi Tech, 28(4): 577-594
Chapter 20
Clustering User Data for User Modelling in

the GUIDE Multi-modal Set-top Box
P.M. Langdon and P. Biswas
20.1 Introduction
The EU GUIDE project is aimed at developing a research-based prototype of an
inclusive and accessible set-top box running on conventional hardware. It utilises
advanced user modelling and simulation in conjunction with a single layer
interface that permits a wide range of input devices and modalities and output
formats and modes. This paper addresses part of the research behind the
development of an advanced user model in order to develop the software
framework, namely the user profile clustering. A range of user centred design
techniques, including focus group, survey and several iterative stages of design
trials have been used overall to create the requirements specification for the
technology framework under development. However, a key module is the inbuilt
user model that allows the system to classify users on the basis of their impairment
level, user interface (UI) behaviour, preferences and context. These profiles are
pre-generated using an advanced cognitive, perceptual and movement simulation,
using parameters clustered from the actual user data. This chapter examines the
contextual background literature, briefly describes the inclusive user centred design
process and shows how this process generated the data necessary for clustering for
cognitive, perceptual and motor-impaired user modelling.
20.1.1 Background and Motivation

GUIDE primarily addresses end users who have mild to moderate impairments.
This target group is identified within the context of literature concerning inclusive
design, accessibility and the digital barriers excluding older technology users from
access to digital technology. An inclusive approach considers the visual, hearing
and touch perceptual capabilities of users in combination and attempts to quantify
capability variation. This approach is well suited to multimodal interface design

196 Langdon and Biswas
where capability impairment in one modality may be compensated for using other
modalities (e.g. hearing impaired users may use visual captioning, avatars and sign
language) or where multiple impairments can be addressed using the performance
gains arising from multimodal interfaces. For instance, gesture, speech and face
recognition for input can be used with sound, touch and graphical output displays
to enhanced effect.
20.1.2 User Centred Design and the Design Process

End user requirements for ageing people and those with impairments were
collected using a mixed methods approach based on the advantages of triangulation
of data sources (Langdon et al., 2003; Flick, 2006). In essence, this approach does
not commit to a single source of data or a single data collection approach. Instead
data is collected from multiple approaches, for example: literature review,
quantitative analysis of data from forums, user trials, user surveys, and
questionnaires, qualitative analysis of observational data from user forums or
interviews, video from user trials and usage ethnography (Langdon et al., 2003;
Flick, 2006). In the case of GUIDE, the framing of the project as a design problem
constrains the triangulation and assists it by directing the focus of comparison on
the design of the final interactions between user, system, and technology and usage
context. The particular methods used are shown in Figure 1. The sampling strategy
employed was opportunistic, and stratified, choosing data sources according to
convenience and resource limitations. In particular, much of the work took
advantage of empirical and observational trials. However, the combination of
multiple sources permits triangulation and thus increased validity and reliability of
qualitative findings (Miles and Huberman, 1994).
Figure 20.1. Schematic representation of development of requirements
We focus on the lower region of Figure 20.1, on the data sources and processes
of clustering user data from: objective performance data from user trials; user
screening survey data; existing data from input device validation trials. The
clustering of data from these sources was intended to define groupings that could
be traced by membership to specific capability levels in specific modalities.
Clustering User Data for User Modelling 197
20.1.3 User Centred Design and the Design Process

The User Model in GUIDE runtime system is incorporated into the framework as a
set of pre-calculated user profiles based on the user data clusterings mapping user
characteristics. It is developed using the modelling (Benyon and Murray, 1993) but
does not include the detailed perception, cognition and motor process models. It
consists of the following two parts:
• A set of sample profiles developed using the clusters described here. The
user profile will be described in a machine-readable and human-readable
format in accordance to a standard set of variables.
• A table mapping user characteristics to interface and adaptation parameters
developed offline using the simulator and user trials.
To address the limitations of existing user modelling systems, we have
developed a simulator (Biswas et al., 2011). The Environment model contains a
representation of an application and context of use. It consists of the Application
model containing a representation of interface layout and application states for the
current task undertaken by a user that will be simulated by breaking it up into a set
of simple atomic tasks. The Context model represents the context of use such as
background noise, illumination and so on and a Device model decides the type of
input and output devices to be used by a particular user and sets parameters for an
interface.
Figure 20.2. Architecture of the simulator
The User model simulates the interaction patterns of users for undertaking a
task analysed by the task model under the configuration set by the interface model.
It uses the sequence of phases defined by the GOMS Model Human Processor
(Card, Moran, Newell, 1983).
• the perception model simulates the visual perception of interface objects. It
is based on the theories of visual attention;
• the cognitive model determines an action to accomplish the current task. It
is more detailed than a GOMS model but not as complex as other cognitive
architectures;
• the motor behaviour model predicts the completion time and possible
interaction patterns for performing that action. It is based on statistical
analysis of screen navigation paths of disabled users.
The details about users are stored in xml format in the user profile following an
explicit ontology.
20.2 Data Origins and Clustering Approach

There were several phases to the data clustering approach:
• obtain and collate survey data and user trial data;
• cluster the survey data by modality: Vision, Hearing, Cognition, Physical;
• reduce the dimensionality of the data set by eliminating non-significant
variables in a k-means clustering;
• take the resulting clusterings and characterise the cluster centres in terms of
the combined contributions to the clusters;
• repeat for User trial data.
20.2.1 Sampling Strategy

For the purpose of devising a working sampling strategy for early pilot trials it was
necessary to adopt a stratified sampling strategy, screening participants who took
part in the GUIDE user survey and allocating them to age and capability ranges of
interest and that were required for analysis. Initially participants were screened into
groups according to the severity of their perceptual, cognitive and motor
limitations using conventional standardised tests, such as the US 4.2.1 AMA
Uniform Measurement Scales (Figure 20.2). For example, the following
classification is taken from Dementer and Anderson, (2003).
Table 20.1. AMA General ability ranges

The relative proportions of participants in sampling categories were therefore

initially decided by opportunistic sampling. This process was further informed by an
Inclusion and Disability analysis based on exclusion calculations of the proportion of
an exemplar EU population (UK GB) who might be expected to fall into the
categories of high, medium or low impairment, and who consider themselves
disabled in some capability area. These figures were derived using calculations based
on the data from the only known complete data set of capability variation publicly
available through the UK Office of National Statistics (ONS). This representative
national data was from a UK disability survey (Disability Follow-up Survey, DFS,
1997). DFS carried out in 1997 and intended to establish the prevalence and severity
of quality of life problems arising from functional impairments (Grundy et al., 1999).
Table 20.2 exemplifies the result of this approach for vision and hearing alone but
figures were also generated for cognition, and physical movement.
Table 20.2. Population and proportion excluded for each of the age ranges and impairment
levels (hearing and vision only)
TYPE OF SUBTYPE OF LEVEL OF PROPORTION OF PROPO RTION OF
CAPABILITY CAPABILITY IMPAIRMENT DISABLED DISABLED
POPULATION POPULATION
EXCLUDED AT EXCLUDED AT
THIS LEVEL THIS LEVEL
40-60 60-90
Vision High 0.04% 0.21%

Perception
Medium 0.44% 2.5%
Low 1.86% 7.41%
No impairment 3.21% 11.88%
Hearing High 0.09% 0.55%
Medium 0.42% 2.49%
Low 2.28% 10.99%
No impairment 4.1% 18.69%
Finally, the gender balance for the population age ranges and the proportion of
individuals experiencing some impairment for each gender were also calculated
using the exclusion estimation from the UK inclusive design exclusion calculator
(IDtoolkit, 1999; Waller et al., 2010).
Because the sampling strategy of older and impaired users in the age groups of
interest was, by necessity, opportunistic, these calculations were used as a broad
indication of the relative sizes of the samples required in each category. This
meant, for example, that roughly twice as many individuals would be sampled in
the older age group compared with the younger age group and that the sample
should be biased towards female rather than male participants. In the actual
sampling these proportions were only partially achieved, the predominant
population being older female, although the age group sampling was partially
successful. The extent which this reflects the sample from Northern Spain was
unknown but comparisons with other populations in the UK and Germany will be
available in the second iteration of user trials to come.
20.2.2 Survey Data Analysis

The full Stage 1 survey data set contained 46 users and a very large number of
variables. The full set is too large to be illustrated. The variables were then
subjected to k-means cluster analysis for each modality, again because of the large
number of variables, and this subsequently allowed the removal of variables that
did not significantly contribute to the high, medium and low clustering. The trial
data contained a prohibitively large number of missing data points for analysis.
20.2.2.1 K-Means Clustering

After collecting survey data from them we have defined the profiles of low,
medium and high levels of severity. In defining the profiles we have used k-means
clustering with a value of k =3. K-means clustering algorithm works like an
Expectation Maximisation (EM) algorithm, which partitions n observations into k
clusters in which each observation belongs to the cluster with the nearest mean.
Each cluster centre defined a profile and only the significant variables (p<0.01)
were used to define the profiles.
20.2.2.2 Removal of Non-significant Variables

Removal of the variables that did not significantly contribute to the high, medium
or low clusters gave the following set. This set of variables appears to be consistent
with literature and has face and completeness validity.
Table 20.3. Final significant variables for each modality
SURVEY DATA REMOVING NON-SIGNIFICANT VARIABLES
VISION HEARING COGNITION MOTOR

Close vision: able to Able to hear a sound of 500Hz TMT test of Diagnosed mobility
read perfectly close to cognitive function impairment
Diagnosed visual Able to hear a sound of Amount of
impairment 2000Hz muscular
weakness
General eyesight Able to distinguish a Able to write
conversation from a noisy
back ground
Able to see at night Amount of limb
tingling
Amount of limb
rigidity
20.2.3 Characterising Cluster Centres

The following section uses the result of the k-means clustering to characterise what
a typical member of a high medium or low impairment group would be capable of,
based on the calculated cluster centres. The profiles of users will ultimately be
formed of a combination of all modalities simultaneously such that a specific
grouping may represent capability levels for users perceptual, cognitive and motor
capability ranges. Clearly, a particular user may present, for example, low vision,
medium cognition and high motor capability levels and it is this that will be
initially used to index the GUIDE user profiles following a successful initialisation
of the application. This may entail a short visual acuity, cognition, physical or
hearing test, requested by GUIDE. Following this, adaption may utilise more
accurate quantitative data to further locate the user’s profile.
20.2.3.1 Cognitive Tests

Before presenting the clustering result it is necessary to describe the cognitive
measures used in more detail. Measures for perception and movement are
essentially self-explanatory. However, cognitive measures were less obvious.
AVLT: The Auditory verbal learning test (AVLT) consists of 15 words which
have to be learned during 5 trials. After every trial, the respondent is asked to recall
as many words as possible from the list. Two measures were used in GUIDE:
1. AVTL 1: this is the short term memory required to recall as many

words as possible.
2. AVLT 5 - 1: this is the learning potential variable. The results of this
test are interpreted taking into account the age of the respondent:
• <74 years old: the expected average on the first trial is 4;
the expected average on the AVLT 5-1 is 5
• 75-84 years old: the expected average on the first trial is 3;
• > 84 years old: the expected average on the first trial is 3;
DST: The Digit-symbol test (from WAIS): It measures information processing
speed (Weschler et al., 1999), requiring rapid matching of learnt digit to symbol
pairings.
• 55-69 years old: 36-43 symbols in 2 minutes expected
• >70 years old: 18-22 symbols in 2 minutes expected
TMT: The Trail Making Test (TMT) measures Cognitive executive function.
This can be interpreted as the ability to use working memory and the number of
tasks that can be kept in mind and dealt with at once. The measure is the time
needed to carry out the task of connecting the drawn circles together in numerical
or symbolic order, with a pencil. (TMT, 1944).
• Less than 50 seconds: no impairment
• 50-60: low impairment
• 60-70 : = medium impairment

• > 70: = high impairment.
However, the normalised test literature occasionally conflicted with the
clusterings resulting from the k-means approach. In some tests some of the scores
interpreted as a level of impairment in the clustering table were not indicated as
impairment in the test rubric. This was assumed to be attributable to differences in
sampling and inevitable inaccuracies of clustering with small samples.
20.2.3.2 K-Means Cluster Centres

The following tables show what levels of the variables characterise the cluster
centres for each modality.
Table 20.4. K-Means cluster centres for the visual and hearing variables
Vision – Cluster Centres

LOW MEDIUM HIGH
Close Vision: able to read until line 20/20 20/60 20/80
Distant_Vision: able to read until line 5/5 5/5 5/20
General eyesight good excellent normal
Seeing at distance good poor poor
Seeing at night normal poor poor
Colour perception good bad bad
Hearing – Cluster Centres

LOW MEDIUM HIGH
Able to hear a sound of 500Hz? Yes Yes No
Able to hear a sound of 1Khz? Yes Yes Yes
Able to hear a sound of 4Khz? Yes Yes No
Able to hear a sound of 8Khz? Yes No No
How do you define your hearing..? excellent good poor
Hear conversation from a noisy
background excellent normal normal
Hear movie dialogue excellent good poor
Hear ringing noises excellent good normal
Hear phone ringing in a movie excellent good poor
Table 20.5. K-Means cluster centres for the cognitive variables
Cognition – Cluster Centres

LOW MEDIUM HIGH
TMT seconds 30 49 136
AVLT series 1 6 5 3
AVLT series 2 10 7 5
AVLT learning potential 7 5 3
WAIS - digit-symbol test 75 30 20
Table 20.6. K-Means cluster centres for the motor variables
Motor – Cluster Centres

LOW MEDIUM HIGH
Hernia,slipped-
Diagnosed mobility impairment no disc no
Amount of muscular weakness never occasional quite a lot
Able to write No difficulty No difficulty Mild
Able to push a heavy door No difficulty No difficulty Mild
Able to change a bulb No difficulty No difficulty Mild
Use of transport (bus, etc.) No difficulty No difficulty Moderate
Amount limb tingling No difficulty Mild Mild
Amount limb weakness No difficulty Mild Moderate
Amount limb rigidity No difficulty Mild Moderate
20.3 Discussion
This chapter has described the context, antecedents and preliminary results of the
GUIDE cluster analysis from a user-centred custom-designed data collection. A
data snapshot of survey data was analysed for highly correlated variables and on
this basis single variables with utility were chosen and used to represent vision,
hearing, cognition and physical movement capabilities. Initial results suggest that
the simple k-means clustering approach yields high, medium and low impairment
cluster centres that can be characterised on the basis of the significant variables
that contribute to them. Both the characterisations and clusterings present face
validity and completeness validity, but this initial data set was smaller than the
expected final data set. However, we conclude that such a clustering is, in
principle, capable of use in the GUIDE framework when combined and indexed
within the GUIDE user model. The result will be that the user interface presented
to the user will be tailored to their perception, cognition and movement capabilities
and that sound, visual display and other outputs will be combined multimodally.
Future clustering could take advantage of quantitative parameters of clustering
variables to position users in the profile space. Clusterings for the survey data (n =
46) were consistent also with expected distributions based on the users sampled
and age estimates. Further work will continue to enlarge the data sets and employ
advanced clustering techniques, such as fuzzy and overlapping clusters, to group
both the survey and trial data as candidate sets for user profiling.
20.6 References
Benyon D, Murray D (1993) Applying user modeling to human computer interaction design.
Artificial Intelligence Review, 7(3/4): 199-225
Biswas P, Robinson P, Langdon PM (2011) Designing inclusive interfaces through user
modelling and simulation. International Journal of Human Computer Interaction. Taylor
& Francis, 1044-7318
Card SK, Moran TP, Newell A (1983) The psychology of human-computer interaction.
Lawrence Erlbaum Associates Publishers, Hillsdale, NJ, US
Demeter SL, Andersson GBJ (2003) Disability evaluation, 2nd edn. Mosby, St Louis, MO,
US
Flick U (2006) An introduction to qualitative research, 3rd edn. Sage Publications Ltd,
London, UK
Grundy E, Ahlburg D, Ali M, Breeze E, Sloggett A (1999) Disability in Great Britain:
Results from the 1996/97 disability follow-up to the family resources survey. Technical
Report 94. Department of Social Security, Leeds, UK
Inclusive Design Toolkit. Available at: http://www.inclusivedesigntoolkit.com/betterdesign/
downloads/exclusioncalc.html (Accessed 20 November 2011)
Langdon PM, Aurisicchio M, Clarkson PJ, Wallace KM (2003) An integrated ethnographic
and empirical methodology in a study of knowledge searches in aerospace design. In:
Proceedings of the 14th International Conference on Engineering Design (ICED’03),
Stockholm, Sweden
Miles MB, Huberman AM (1994) Qualitative data analysis. Sage Publications Ltd, London,
UK
TMT (1944) Army individual test battery. Manual of directions and scoring. War
Department, Adjuvant General’s Office, Washington, DC, US
Waller SD, Williams EY, Langdon PM, Clarkson PJ (2010) Quantifying exclusion for tasks
related to product interaction. In: Langdon PM, Clarkson PJ, Robinson P (eds.) Designing
Inclusive Interactions. Springer, London, UK
Wechsler D (1999) Adaptación española de la Wechsler Adult Intelligence Scale-III
(WAIS). TEA, Ediciones, SA, Madrid
Part V
Designing Inclusive
Architecture
Chapter 21
Inclusive Built Heritage as a Matter of

Concern: A Field Experiment
A. Heylighen
21.1 Introduction
Europe’s built heritage is the world’s most diverse and rich patrimony, and an
important component of individual and collective identity. Its societal relevance is
inextricably linked to sustainability: by opening up built heritage and using it
appropriately, its upkeep is best secured and its protection from decline guaranteed
(Adriaenssens et al., 1998; Gobyn and Knops, 2000). Integrated conservation
therefore strives to give built heritage a contemporary role in society. At the same
time, inclusion policy strives for universal participation in society, which requires
that environments can be reached, entered, interpreted and used by people with
diverse and changing abilities.
When built heritage plays a contemporary role in society, and different people
participate in society, both meet. Making built heritage inclusive - i.e. reachable,
accessible, understandable and usable for as many people as possible - is a highly
complex matter, however. Proposals to make historic buildings more inclusive tend
to raise objections from conservation authorities, which guard the historic values of
built heritage. Current approaches to accessibility do not seem to deal with these
concerns well. Particularly telling in this respect is the accessibility legislation for
public buildings recently issued in the region of Flanders: monuments that are
provisionally or definitely protected, or building sites located in (provisional or
definite) conservation areas, are exempt from its application. In this relatively small
region, this comes down to no less than 10,000 buildings listed as protected. Built
heritage thus remains out of reach, both practically - from the perspective of disabled
people - and legally - in terms of building regulation.
Is making built heritage (more) inclusive really beyond reach, the proverbial
exception to the rule? Or can we address it from a different angle? This paper reports
a field experiment that addresses the inclusivity of built heritage in a different way,
by allowing a group of people to become concerned with this issue. The context of
the field experiment is a European university, whose campus features a considerable
number of protected buildings. After introducing the origins and set-up of the

208 Heylighen
experiment, we report what insights it has yielded so far, how these are received by
different people and how they impact the real-world situation on campus. We draw
on field notes and pictures taken during participation in building visits, analysis
reports, and written and oral accounts by different people involved in or affected by
the experiment. After comparing the approach explored with more traditional ways
of addressing the inclusivity of built heritage, we conclude by questioning its
transferability to other real-world contexts.
21.2 Context
A university is a very specific and in a sense ‘unique’ institution (Biesta and Simons,
2009) in that it is multiple (id)entities at the same time (Heylighen and Nijs, 2011):
an institution of higher education where people study; a research institution where
people conduct research; a built environment featuring campuses, buildings and
rooms that accommodate students and staff; and an organisation with several
departments, including technical services that take care of the accommodation. These
multiple identities make a university an ideal setting for a field experiment.
A university is first and foremost an educational setting where people study.
Since young people are educated as much by example as through teaching,
environments that segregate teach acceptance of segregation, whereas inclusive
environments teach inclusion (Welch and Jones, 2001).
To some extent, a university is also a miniature version of society. Several agents
involved in or affected by making built heritage more inclusive are present in its
organisation: the building owners, architects and other built environment
professionals (e.g. building conservation specialists), services for construction,
management and maintenance, and building users, i.e. students, staff, and visitors,
both young and moving into old age, with and without disabilities. In line with the
exigencies of a ‘real’ experiment, we can thus say that the university offers an
‘ecologically valid’ setting.
Of all those active at a university, students, staff and visitors living with diverse
abilities and conditions can be considered as “user/experts”, a term introduced by
Elaine Ostroff (1997) to denote “anyone who has developed natural experience in
dealing with the challenges of our built environment”. Their experiences may offer
designers unique and expanded insights - see for instance Pullin (2009) and Helen
Hamlyn Centre (2009), yet in relation to built heritage their voices often remain
silent. Built heritage is typically approached in an essentialist way that focuses on the
built environment in itself. Giving voice to disabled building users in improving its
inclusivity, however, shifts the focus from built heritage to how people experience it.
Such a relational approach resonates with social conceptions of disability which,
unlike prevailing medical conceptions, place the body in its sociomaterial context,
recognising the interplay between physiological condition and features of the society
one lives in (Butler and Bowlby, 1997). In relation to built heritage, this move to
embrace disability as a social issue can be traced in the strategic framework for
access to historic and heritage buildings developed by English Heritage (Adams and
Foster, 2004). It is this framework which inspired us to conduct a field experiment at
Inclusive Built Heritage as a Matter of Concern 209
our university to involve disabled students, staff and visitors in addressing the
inclusivity of built heritage on campus.
21.3 Origins and Set-up

The idea of conducting the field experiment arises at a point when the university, in
its role as building owner, decides to obtain expert advice from an official
accessibility office on the inclusivity of its built patrimony. By way of trial one
protected university building is subjected to an accessibility audit. The audit is
performed by a professional accessibility advisor - a specialised architect - who
assesses the building using a standardised checklist, and formulates a proposal to
address the problems identified in a phased way.
Rationale. The approach adopted in the field experiment is intended not as an
alternative for, but rather as complementary to professional approaches like the
accessibility audit. Instead of imposing certain solutions upon architects, it aims to
inform them or at most make suggestions to them, offering a basis for making design
decisions while leaving the actual design up to them.
Buildings. The buildings considered in the field experiment are protected
buildings on campus. They are selected in consultation with architects of the
university’s technical services. Preference goes to buildings for which works are
planned in the near future. So far three buildings have been addressed: the Van
Dalecollege, a 16th century college accommodating the university’s student services
and student housing; the Arenbergcastle, a 16th century building housing the
architecture department; and the Pauscollege, a late 18th century college used as a
dorm for 180 students plus a branch of the university restaurant. In an early version,
the approach explored in the experiment was also applied to the Grote Aula, a 19th
century auditorium used for lectures and music events.
Teams. Each building is analysed by five teams. Every team is composed of one
user/expert and two Master students in architecture (or, in some cases, one student
and one researcher). User/experts include students, staff and visitors with a mobility
impairment (using a wheelchair, having difficulty walking), a sensory impairment
(blindness, low vision), of a diagnosis on the autistic spectrum. The architecture
students attend an elective course on inclusive design.
Teams visit the building being considered and identify its qualities and
weaknesses from the perspective of the user/expert in the team. Based on an earlier
experiment with professional architects, not related to built heritage (Heylighen et al.,
2009), we expected that during these visits a particular dialogue would develop
between the user/expert on the one hand and the architecture students on the other
hand: a dialogue that is embodied in nature, unfolds in situ, and involves a particular
knowledge transfer (Heylighen and Nijs, 2011). Through such a dialogue, experience
is being framed: both the user/expert and the (student) architect find themselves in a
considering stance - considering their experience of the building for the former,
considering design practice for the latter.
Output. The architecture students write an analysis report summarising the major
insights gained during the visit of their team. The report is not normative in that it
210 Heylighen
informs us about how the user/expert in the team experiences the building visited,
rather than prescribing what should be altered. It is narrative in that it addresses the
building’s spatial qualities and obstacles in a way that respects the intricate
relatedness of things in how the user/expert experiences it, rather than point-by-point
(as in say a standardised checklist). The report is documented with photos and
graphic material that resonate with architects’ visual way of working.
Reports are shared and discussed with the other teams analysing the same
building, thus augmenting their validity, and with architects and other built
environment professionals of the technical services.
21.4 Framing Experience, Nuancing Inclusivity

The analysis reports offer a highly nuanced picture of the inclusivity of the university
buildings visited, revealing issues that may easily be overlooked in making built
heritage more inclusive or that built environment professionals may not be attuned to.
Moreover, besides unforeseen issues to address, they also point to unforeseen
opportunities for improvement.
21.4.1 Different Needs or Different Reasons

The teams’ analysis reports of the building visits show how different the needs put
forward by different people - or even by one person - can be. A user/expert who has
difficulty walking, sometimes uses a wheelchair, but visits two university buildings
(Arenbergcastle and Pauscollege) on foot. During these visits, he sometimes points
at aspects that do not raise a problem at this point, but would if he were using his
wheelchair; or vice versa. For example, slopes are very handy when in his
wheelchair, but on foot he prefers a well dimensioned staircase to a slope, because on
the latter he has more difficulty keeping his balance. Inclusivity reveals itself here not
as a timeless and invariant feature (as it does in traditional approaches), but instead
shows its ambiguity and situational character.
However, in addition to the differences in (and at times contradictions between)
the needs pointed out by the user/experts, the analysis reports also reveal building
aspects that are experienced as problematic by several of them. For example, the Van
Dalecollege, Arenbergcastle and Pauscollege each have a courtyard covered with
cobblestones. Their unevenness causes problems for several user/experts: for the
person having difficulty walking it makes using a cane more difficult and increases
the risk of stumbling; for the wheelchair users, it provides a bumpy ride; and for the
blind participants it makes walking with a white cane difficult - cobblestones lying in
the same direction cannot be felt as a guiding line.
Similarly, several user/experts complain about the (lack of) light in the
Pauscollege. The architecture students collaborating with a person with low vision
notice that the transition from dark to light(er) spaces - and the other way around -
constitutes a considerable threshold for her. As her eyes need to adjust, she walks less
swiftly and less spontaneously through the building. The recognisability of building
elements diminishes considerably in dark spaces. For the architecture students it does
not make a difference in which corridor they are walking, yet they feel that the
user/expert walks more cautiously through the darker corridors. A person with autism
also mentions the lack of (sufficient) natural light in the Pauscollege, which he
considers especially problematic in the long windowless corridors along the student
rooms. Deprived of contact with outside, he does not know on which side of the
building he finds himself. It is interesting also that the only corridor he finds beautiful
does have sufficient light. This enables him to see better how the space is finished.
More generally, several user/experts characterise the Pauscollege as unpleasant.
The blind person finds the building “not cosy at all.” It is “way too big” and there is
“not much order”. The user/expert with autism has the impression that “lumber is
lying everywhere” which he finds disturbing. After the visit, he is happy to be
outside again because he dislikes the interior of the building and has an oppressive
feeling inside. Asked what he finds unpleasant, he refers to the “prison corridor”.
This specific corridor is more spacious than the other ones, but because the “prison
feeling” prevails, he finds it particularly unpleasant.
In these examples some aspects of buildings are pointed out as problematic by
several user/experts, be it for different reasons. By not merely identifying problems,
but explicitly describing the different reasons cited for them by the user/experts, the
analysis reports can offer architects a better basis for designing more inclusive
solutions. Standard or conventional formulations of problems tend to trigger standard
or conventional solutions. By contrast, the descriptions of the reasons for these
problems in different wordings (e.g. “cosy”, “lumber lying everywhere”, a vague “too
big”) or metaphors (“prison corridor”, “prison feeling”) leave designers more
degrees of freedom through their semantic openness.
Still in the Pauscollege, another space the user/expert with autism finds beautiful
is the hall with the old staircase. The hall is light and spacious, and the rustic wood
offers a beautiful contrast with the white painted walls. More generally, old staircases
in the different buildings are clearly appreciated by several participants. In the
Arenbergcastle, for instance, a staircase in the porter’s lodge is praised in several
analysis reports for its comfortable dimensions and its handrail. The handrail does
not only offer a good grip, its banisters make the staircase clearly recognisable as
such to a blind user/expert. Also in the Pauscollege one particular staircase is
described as very comfortable and its handrail as offering good grip. These examples
demonstrate that the analysis reports include building aspects that are valued by
several user/experts, and not only those that are criticised by them. This enables
architects designing inclusive solutions to build on strength instead of focusing on
faults and weaknesses only.
In summary, then, the approach adopted in the field experiment shows that needs
may differ considerably depending on the person or situation, thereby unmasking
inclusivity of built heritage as ambiguous and situational. For some aspects of
buildings, however, several user/experts agree that they are either problematic or
valuable. This occasional convergence between different user/experts might suggest
a certain ‘universality’ of the problem or solution considered, yet universality or
generalisation is not an aim in itself here. More interesting is the fact that architects
are offered insight in why user/experts either dislike or value certain building aspects.
212 Heylighen
21.4.2 Use and Organisation

When visiting the Van Dalecollege, two visually impaired user/experts point at the
lack of clear organisation (Heylighen et al., 2010). A person with low vision has the
impression that the building complex is not designed as a whole. He finds that having
to search for a room is not very user-friendly, and increases the importance of
inclusive signage. A blind person, for her part, finds the building inconveniently
arranged. For her to use it independently would require a clear explanation of its
appearance, location, orientation and structure. Nevertheless, she is able to find the
reception by herself because it is near the entrance, where she would look intuitively
(and because of its smell, reminding her of a library or journals). The arrangement of
the ground level seems relatively convenient to her, yet overall, she characterises the
building as a true labyrinth requiring supreme concentration to navigate. A person
with autism also has trouble with the lack of clear organisation. Except in places he is
familiar with (because he has been there before), it is difficult for him to locate at
which point in the building he finds himself.
When entering the main entrance of the Arenbergcastle, a blind user/expert is
relatively quickly on to the fact that the building is structured around a courtyard. He
derives this from what he hears. He describes the covered entrance as a passageway
in between two buildings (he notices an echo), which is followed by an open space.
The rectangular shape of the courtyard makes it easy to orientate himself. The
secretariat of the department housed by the building, however, is located in an
illogical spot: while he would expect it close to the main entrance, it is located in a
side wing.
A user/expert having difficulty walking suggests changes to the castle’s
organisation as well. The entrance to the porter’s lodge would be much more
accessible to him when using the back door instead of the door giving on to the
courtyard. The same applies for the seminar rooms. By recognising the entrance via
the current secretariat as a main entrance to the seminar rooms, people are not
obliged to cross the bumpy cobblestones in the courtyard.
For a user/expert with autism, the experience of a space seems to be influenced
considerably by its use. The entrance hall in the Pauscollege, for instance, has large
windows which let in a lot of light. Still he does not find it a pleasant place to wait, as
the noises of the drinks machine and of the people passing are too disturbing.
Similarly, the big spaces in the Arenbergcastle used by architecture students as
design studios probably would not be very suitable for him to work. These spaces
may be very busy, with students and staff running in and out. When entering the
room you are directly confronted by the people present. For him, he says, the design
studios perhaps would be better subdivided in smaller, structured spaces that are
more or less separated from each other in terms of view and sound.
These examples illustrate how the approach adopted in the field experiment
explores and evaluates built heritage as a physical entity, but also considers how it
works. The problems experienced by the user/experts turn out to be caused not only
by material barriers raised by the historic building itself; major problems - and thus
also possible solutions - relate to how the building is used and how this use is
organised. This attention to use and variation in use resonates with architects’ core
business: the organisation of space rather than the physical building as such.
21.4.3 New Weaknesses, Old Strengths

In the Arenbergcastle, recently introduced building elements turn out to cause major
problems to several user/experts. A case in point are the concrete platforms that were
installed a few years ago in front of both entrances to the west wing, and which are
experienced as highly problematic by a blind person, a person with low vision and
one who has difficulty walking. They find the platforms very dangerous because of
the lack of handrails, contrast and marks. The door handle of the new outside door is
not easy to find, it is not recognisable as such and is difficult to grasp. Inside the
castle, the staircases causing most problems are the most recent ones: the spiral
staircase close to the secretariat, and the escalator leading to the staff room above the
secretariat. The spiral staircase is experienced as very unsafe because the steps are
irregular, a handrail is provided on the narrow side of the steps only, it is unstable
and interrupted at two points. Interestingly, user/experts find the old staircases - near
the seminar room and in the porter’s lodge - much more comfortable.
Conversely, historic elements do not always turn out to be problematic; on the
contrary. The different shapes of door handles throughout the castle offer a very
pleasant surprise to a user/expert with low vision. Moreover, they mostly contrast
very well with the door, which makes them clearly visible. She also appreciates the
fact that within the castle and the porter’s lodge the relatively logical structure of
spaces is preserved, so that searching for a room does not take long. The rooms are
rectangular, which facilitates orientation.
A user/expert with autism seems to regret that the original functions are no longer
visible in the way the castle is currently used. The building would be more readable
to him if some relation existed between old functions, e.g. “kitchens” and “salons”,
and new ones. Although he realises that this is impossible as the building now has a
completely different function, it would make the functional organisation better
understandable for him by offering clues on how to orientate himself in the building.
To him, overview and organisation are important in space, but also in time. Overview
and organisation in time assist him in handling his environment and anticipating new
situations.
Together these examples indicate that recent interventions in historic buildings
are not always an improvement in terms of inclusivity; conversely, historic elements
are not always problematic. This suggests that the idea that inclusivity of built
heritage is problematic as a result of its age, and that newer interventions are better, is
undeserved. In other words, by shifting the focus from the historic building to how it
works for people with diverse abilities and conditions, it becomes clear that built
heritage may offer inclusive designers not only a major challenge, but also a source
of inspiration.
21.5 Discussion
Feedback from built environment professionals of the technical services suggests that
they value the analysis reports considerably. As one architect formulates it: “I found
the subjective analyses highly interesting and in many respects they actually taught
214 Heylighen
me more than the objective analysis of the accessibility office. I find the added value
thus very high.” Compared to the accessibility audit conducted by the professional
accessibility advisor, the architect especially seems to appreciate the nuanced
approach adopted in the field experiment: “An important aspect is the broadening of
the term accessibility by including very diverse disabilities, also and above all those
whereby the person is not “entirely” blind or chained to the wheelchair. The
experience of the person with autism surprised me in the most positive sense: he
uncovered in a very direct way problems (…) which we all do sense but never can
point to that well.”
Compared to the professional accessibility audit, the field experiment approaches
inclusivity of built heritage on campus from a completely different angle. To some
extent, inclusivity of the built environment can be - and often is - considered as what
Bruno Latour (2005) refers to as a matter of fact. Similar to AIDS, poverty, global
warming and equality, it is often something we are detached from, taken care of by
state officials or experts, instead of something to which we, as a public, are exposed
or attached (Simons and Masschelein, 2009). Accessibility legislation translates
inclusivity into facts (or indicators and averages) by fixing maximum heights of
thresholds and minimum widths of doors, which in turn can be objectively measured
by professional accessibility advisors performing accessibility audits. Reducing
inclusivity to the realm of matters of fact, in which accuracy becomes the closing
argument of professional experts, leaves those affected by it - the disabled people
themselves - seemingly incapable of joining the dialogue because they are supposedly
no experts in the field (Heylighen and Nijs, 2011).
As pointed out in the introduction, however, inclusivity of built heritage is often
not covered by accessibility legislation, or is considered as an exception to the rule.
The approach adopted in the field experiment acknowledges that there is hardly any
regulation available to address inclusivity of built heritage (including several
university buildings), and that the traditional specialisation and available expertise is
inadequate to solve this problem. Therefore, the approach allows for a group of people
to become concerned with or attached to this issue - architecture students, disabled
students, staff and visitors, and staff of the university’s technical services. In other
words, inclusivity of the built environment is not presented as a matter of fact. Rather,
through analyses of university buildings in collaboration with user/experts, it is made
perceptible in the public sphere and gradually becomes a matter of concern (Callon,
2005; Latour, 2005).
Feedback from the architecture students suggests that they experience this
alternative approach as highly motivating and insightful, but also as very unusual and
therefore somewhat confusing. From other courses, they have become used to the fact
that teachers have the necessary expertise to offer the (or at least a) solution, and they
are surprised to discover that for this issue, this is not the case. In the real-world
situation on campus, inclusivity of the built environment presents itself to the students
as ambiguous and situational. They learn to be affected in new ways by the same
issues (Latour, 2004; Despret, 2004). Through the particular dialogue with
user/experts in situ, i.e. in the protected buildings under consideration, the attention of
the architecture students is being trained.
One user/expert attests to this learning process: “[the student] was open to it. I
could clearly notice that by talking to her.” The user/expert enjoyed participating in
the building visits, and would like the approach to become a compulsory topic of the
students’ program, allowing more people to become attached to the issue: “As an
elective course, you only have motivated students. That’s of course an advantage. But
then it’s something that is possible but not compulsory. (…) The more people you
address, the more obvious it probably becomes. Maybe it’s good to strive for a
matter of course instead of a possibility.”
Interesting to notice, however, is that even as part of an elective course, the
approach has a major impact on local decision making. The insights gained through
the visits meanwhile have motivated and enabled the technical services to implement
major alterations in some of the buildings visited. As we write, the Grote Aula is
undergoing major interventions to improve its acoustic comfort, which are directly
motivated by insights gained through the analyses with user/experts. The outcome of
these also played a crucial role in the negotiations with and convincing of the
conservation authorities. For the Van Dalecollege, input from the user/experts
unlocked the impasse the student services had ended up in (Heylighen et al. 2010).
The lack of organisation pointed out by several user/experts inspired major
organisational interventions to rearrange the student services more logically in the
available space so that all students can consult them, and yet interventions which
require altering the historic fabric remain limited (ibid.). Interestingly, these
organisational interventions come down, to a large extent, to restoring the logic
present in the original building, making use of its inherent qualities. By shifting the
focus from the protected buildings themselves to how people experience them, it
becomes clear that improving their inclusivity does not necessarily require
supplementing the present situation with new (material) layers; leveraging concepts
already present in historic layers may improve the value of the buildings from an
inclusive perspective, while respecting their heritage value.
21.6 Conclusions
Is making built heritage (more) inclusive the proverbial exception to the rule? Or can
we address it from a different angle? Starting from the real-world situation on a
university campus, this paper has demonstrated that it is possible to approach
inclusivity of built heritage in a different way. Key to the approach adopted in the
field experiment is that it allows for a group of people to become involved with or
attached to this complex issue as a matter of concern, rather than considering it as a
matter of fact. While the approach originally was not intended to be political, it turns
out to have a considerable impact on local decision making, which in turn impacts
the inclusivity of built heritage on campus. Concerns do matter - and are (cap)able of
mattering - apparently! Therefore it is tempting to suggest transferring the approach
to other real-world contexts. Yet we should keep in mind that the field experiment
took place in the ‘unique’ context of a university. In order to investigate to what
extent the approach is transferable to other real-world contexts, we seek to extend the
field experiment to protected buildings off campus. In addition, it would be
interesting to investigate how the skills developed by the architecture students
involved in the experiment are received in their professional situation after
graduation.
216 Heylighen
This research has received funding from the European Research Council under the
European Community’s Seventh Framework Programme (FP7/2007-2013)/ERC grant
agreement No. 201673. The author thanks all who contributed to the field experiment,
in particular the user/experts, architecture students, researchers, and staff of the
technical services. Special thanks go to Stijn Baumers, Jasmien Herssens, Sam
Michiels, Greg Nijs and Iris Van Steenwinkel for their invaluable comments on earlier
versions of this paper.
21.8 References
Adams J, Foster L (2004) Easy access to historic buildings. English Heritage, Swindon, UK
Adriaenssens I, Hendryckx M, Van Alsenoy J (1998) Het geheim achter een open deur [The
secret behind an open door]. Koning Boudewijnstichting, Brussels, Belgium
Biesta G, Simons M (2009) Higher education and European citizenship as a matter of public
concern. European Educational Research Journal, 8(2): 142-145
Butler R, Bowlby S (1997) Bodies and spaces: An exploration of disabled people’s
experiences of public space. Environmental and planning D: Society and Space, 15(4): 411-
433
Callon M (2005) Disabled persons of all countries, unite! In: Latour B, Weibel P (eds.) Making
things public: Atmospheres of democracy. MIT Press, Cambridge, MA, US
Despret V (2004) The body we care for. Body and Society, 10(2-3): 111-134
Gobyn R, Knops G (2000) Erfgoedzorg in de 21ste eeuw [Built heritage care in the 21st
Century]. Koning Boudewijnstichting, Ghent, Belgium
Helen Hamlyn Centre (2009) DBA inclusive design challenge. Helen Hamlyn Centre, London,
UK
Heylighen A, Herssens J, Froyen H (2009). Architecture criticism blindfolded. In: Proceedings
of International Conference on Inclusive Design (INCLUDE 2009), Helen Hamlyn Centre,
London, UK
Heylighen A, Neyt E, Baumers S, Herssens J, Vermeersch P-W (2010) Conservation meets
inclusion. Model meets reality. In: Proceedings of the 5th Cambridge Workshop on
Universal Access and Assistive Technology (CWUAAT), Cambridge, UK
Heylighen A, Nijs G (2011) Studying (architecture) in dialogue with disability. In: Simons M,
Decuypere M, Vlieghe J, Masschelein J (eds.) Curating the European University. Leuven
University Press, Leuven, Belgium
Latour B (2004) How to talk about the body? Body and Society, 10(2-3): 205-229
Latour B (2005) From realpolitik to dingpolitik or how to make things public. In: Latour B,
Weibel P (eds.) Making things public: Atmospheres of democracy. MIT Press, Cambridge,
MA, US
Ostroff E (1997) Mining our natural resources: The user as expert. Innovation, Industrial
Designers Society of America (IDSA), 16(1): 33
Simons M, Masschelein J (2009). The public and its university: beyond learning for civic
employability? European Educational Research Journal, 8(2): 204-217
Welch P, Jones S (2001) Advances in universal design education in the United States. In:
Preiser WFE, Ostroff E (eds.) Universal Design Handbook. McGraw Hill Professional, NY,
US
Chapter 22
Designing a Virtual Environment

Framework for Improving Guidance for the
Visually Impaired
S. Kammoun, M.J-M. Macé, B. Oriola and
C. Jouffrais
22.1 Introduction
Electronic Orientation Aids are dedicated to orientation assistance for the visually
impaired. They are made of at least 3 essential components: 1) A positioning system
(e.g. GPS); 2) A Geographical Information System (GIS) that includes both a
digitised map and a software designed to select routes, track the traveller’s path, and
provide him with navigation information; 3) A User Interface (UI) that relies on non-
visual (usually auditory or tactile) interaction.
These three components could all be the cause of usability issues. The first major
issue is error in GPS positioning that is frequently greater than 20 metres (especially
in cities), which is really not compatible with VI pedestrian guidance. Secondly, GIS
usually contain exclusively road networks, and hence lack pedestrian-related
information. Finally, the interaction with an EOA is a key element, and it must be
designed from the beginning for visually impaired users on the move. Virtual
interactive environments may represent a valuable platform to selectively isolate GIS
or UI-related issues for being safely and systematically tested in laboratory before
on-site evaluations.
With the rise of the number of EOAs, research groups were interested in virtual
environments (VE) to assist visually impaired people in learning Orientation and
Mobility skills (Sanchez and Tadres, 2010). Those systems have been developed to
help construct a mental representation of space from tactile or auditory cues, to
increase spatial cognitive abilities (Mereu and Kazman, 1996), and to give visually
impaired people a tool for safely exploring and learning about new spaces on their
own (Schloerb et al., 2010). Generally, these systems are designed to allow VI users
to explore virtual representations of real or abstract (e.g. a labyrinth) spaces, as well
as to interact with objects within these spaces (Sanchez and Hassler, 2006). However

218 Kammoun et al.
they never rely on defective positioning, nor on GIS adapted to visually impaired
pedestrians.
With the purpose of improving autonomy in the mobility of the visually impaired,
this study proposes the exploration of a virtual environment (VE), through auditory
and haptic interaction. This platform allows the user to navigate in a virtual
environment representing an existing space (based on the GIS of a city for instance).
The aim of this platform is to systematically test several guidance processes before
implementation in an EOA. These tests will help to determine which guidance
process is the most efficient to compensate for inaccurate GPS positioning, while
improving mobility and orientation at the same time.
In the following section, we describe the guidance process generally used in
EOAs. In section 22.3 we present the proposed virtual environment framework. We
focus on the GIS component, user interfaces and mobility within the VE, and
technical issues during design and implementation steps. Finally we discuss how this
system will help to benchmark different guidance processes used in EOAs, and
evaluate resultant navigation performance and cognitive mapping.
22.2 Guidance in Electronic Orientation Aids

When compared to car navigation, it is obvious that pedestrian navigation, especially
with visually impaired travellers, imposes additional requirements upon the
electronic aids. Studies on human navigation (see e.g. Loomis et al., 2001) report that
there are two distinct methods for keeping track of position and orientation during
travel. The first mode is called landmark-based navigation. In this mode, visual
landmarks provide the traveller with direct feedback regarding current position and
orientation. When considering navigation of visually impaired pedestrians, visual
landmarks are inoperative, but are replaced by auditory, somatosensory or olfactory
landmarks. In the second mode, called path integration, the traveller uses the sense of
body motion (kinaesthetic feedback) to update his current position and orientation
relative to the starting point. This mode is operative in visually impaired pedestrian
navigation, but estimated position rapidly drifts if landmarks for position correction
are too sparse. Then an efficient EOA should provide the visually impaired traveller
with mobility instructions, but also with frequent and usable landmarks that allow
both landmark-based navigation and path integration.
In EOAs, the guidance process consists in first identifying the location of a
visually impaired user relative to the expected trajectory and then providing her/him
with the appropriate direction instructions, or with pertinent information about the
surroundings. Hence we can define guidance process according to three main steps:
1) Route selection procedure for computing the optimal itinerary; 2) User tracking
for estimating his current position; 3) Display of navigation instructions and spatial
descriptions to orientate the traveller and improve his mental representation of the
environment (Kammoun et al., 2011). In order to systematically and efficiently test
different guidance processes, these three modules should be operating at once, which
is very challenging with visually impaired users moving in a real environment. To
palliate this difficulty, we designed a controlled virtual environment for testing
different guidance processes potentially used in EOAs.
Designing a Virtual Environment Framework 219
22.3 A Platform for Improving Visually Impaired

Guidance
The system described here is a research tool aimed at designing and evaluating
improved guidance for the visually impaired. Our objective is to use this platform to
evaluate different guidance processes to be implemented in an EOA. In a second
step, the platform will also serve to study the enhancement of cognitive mapping of
the visually impaired during guidance. Improved guidance is based on the presence
of numerous geolocated pedestrian-related data (pedestrian paths, non-visual
landmarks and points of interest) that are annotated in the GIS database and
displayed during navigation. In the following sub-sections, we focus on the technical
development of the platform.
22.3.1 Adapting GIS Components to Visually Impaired

Needs
Most spatial databases used in GIS have been developed without considering the
needs of the Visually Impaired. Very few studies have aimed to identify their needs
and proposed an annotation of geographical objects that should be included in
specialised GIS databases. An interesting classification has been proposed (Golledge
et al., 1998). They divided pedestrian features into four classes: (1) transportation
(e.g. roads, bike paths, walkways, car parking areas, bike parking), (2) buildings, (3)
land use (e.g. open space, recreation, vegetation), and (4) other objects (e.g. light
poles, telephones, and stairs). This classification is interesting but still misses
important data to select the best adapted route for the VI. In our research group, we
have adopted a long-term user-centered design approach in collaboration with the
Institute for the Young Blind (CESDV-IJA, Toulouse) (see Brock et al., 2010). For
this specific project, we interviewed 19 users to define more precisely their needs as
well as their degree of autonomy and technological knowledge. The target population
comprised 7 females and 12 males with a mean age of 37. For daily mobility, 5 of
them use a guide dog and 10 use the white cane. The 4 remaining prefer to have a
person to guide them. All of them are legally blind and expressed their motivation
and agreement to participate in this project. We had three meetings with four
different O&M instructors from the CESDV-IJA. They precisely described the
different steps and techniques that they teach to blind persons during O&M training.
We also analysed through video and a posteriori interviews the O&M behaviour of
two blind users (one with a white cane and one with a dog). We finally performed
three brainstorming sessions with at least 4 blind users in which we investigated
issues with VI navigation. Taking into consideration these different data, we finally
proposed an annotation of geographical data including three main classes: (1)
Walking areas that compose the pedestrian network (sidewalks, pedestrian crossing,
etc.); (2) Non-visual landmarks corresponding to locations that can be detected by a
VI pedestrian on his/her own (e.g. change in the pavement texture or inclination,
street furniture, odours or sounds, etc.). These landmarks are either decision points
(e.g. when to turn) or confirmation points (e.g. a good choice was operated) points;
220 Kammoun et al.
(3) Points of Interest (POIs) such as places or objects that are potential destinations
(e.g. public building, shop, street furniture like mailbox or bus stop, etc.). When POIs
are not a destination per se but near the itinerary, they may be useful to figure out
how the different elements in a city are spatially organised. In addition, both
landmarks and POIs may subserve landmark-based navigation as well as path
integration.
We used data from Open Street Map (OSM) to construct the GIS database. OSM
is an open source project used by a large and dynamic community. The main
advantage is that OSM is an open resource, and it is easy to add data and features to
the database via a simple editor (JOSM). In addition, the data can be shared online.
Hence it is easy to annotate the OSM database according to the proposed
classification by editing nodes, ways, metadata tags and relations.
22.3.2 Avatar Mobility

We chose to rely on a force feedback joystick to control the exploration, because it is
a convenient device to manipulate several dimensions at once (heading, translations
and rotations), and its sensitivity and gain can be adjusted to fit real human walking
speed. Forward/backward movements of the joystick allowed displacement
respectively forward and backward in the VE. Left/right joystick movements
controlled body rotation angle. In order to minimise the complexity of input
interaction, translation and combination of two movements were not allowed. As the
user moved the joystick, the system tracked the direction of the movement and
continuously updated the position of a corresponding avatar in the VE while at the
same time providing auditory feedback in relation to the displacement (step sounds).
22.3.3 Output User Interface

In order to design the output interface, we organised two separate brainstorming
sessions with expert and novice users of EOAs. Our goal was to define the type
and quantity of information required during guided travel, as well as appropriate
modalities that do not interfere with the learned O&M techniques and abilities.
These two sessions made it clear that during a navigation task, an EOA should
provide two classes of information: (1) Direction instructions, i.e. turn-by-turn
instructions, and (2) Space-related information (landmarks for navigation, but also
information about the surroundings, description of difficult points, etc.). To display
both of them, an adapted interface is required, which must rely on non-visual (e.g.
auditory or somatosensory) modalities. Text To Speech (TTS) as well as binaural
synthesis - which provides virtual 3D sounds at any desired location in the
listener’s space - have been evaluated in real navigation context (Loomis et al.,
1998; Gaunet, 2006) and within a virtual environment for orientation and mobility
training (Sanchez et al., 2009; Schloerb et al., 2010).
In real navigation tasks, the somatosensory modality is another efficient output
modality in different situations, e.g. when natural sounds are critical and should not
be masked, when ambient noise level is too high, or when user impairment prevents
auditory-based interaction (visual impairment is sometimes associated with

auditory impairment). Different studies have shown the usability of tactile displays
for presenting directions by mapping them onto body locations. Information
regarding the expected direction was provided through the activation of a given
vibrator mounted in a belt (Pielot et al., 2008), or within a backpack (Ross and
Blasch, 2000). Haptic feedback has also been used within virtual environments
designed for supporting orientation and mobility training (Schloerb et al., 2010).
In the platform, we implemented a feedback editor that allows the addition of
single or combined feedback on objects (TTS and/or 2D/3D sound on the audio
channel; vibration and/or force on the joystick). We first implemented a footstep
feedback that was related to walking speed. Additional feedbacks allowed the
display of both direction instructions, obstacles encountered, and spatial
configuration information (landmarks, POIs). We suggest that a combination of
TTS and binaural synthesis is an interesting solution for guidance as the two kinds
of information are easily distinguishable. Spatialised audio was mainly used for
directional information. Indeed it allowed the user to hear the direction (Right/Left)
of sounds in the VE as if he were standing at the location of the avatar. Spatialised
TTS was also used to render the location of environmental features (landmarks and
POIs). As binaural estimation of distance is rather limited in humans
(Middlebrooks and Green, 1991), TTS or spatialised TTS seemed advantageous to
render this cue. We also designed force-feedback effects that we can attach to
different textures or obstacles in the VE. Hence the user felt the variations in
ground texture and the presence of steps (e.g. sidewalk) or obstacles (e.g. wall) as
in a real navigation task using the white cane. These somatosensory cues may also
serve as landmarks.
22.3.4 Virtual Environment Implementation

As mentioned above, we used the OSM database as the original source of
geographical information in the GIS database. In order to extract information from
the OSM database, we generated an XML file from a selected area. This XML file
was then parsed to get the geometry and attributes of roads, walking areas,
buildings, landmarks and POIs. We finally converted the Word Geodetic System
coordinates (WGS84) used in OSM database to a Cartesian coordinate system in
order to build the virtual environment. Three important classes of objects were
included: 1) Polygons that represent the different buildings, walls and car park
areas. Each building was defined by its name or function, extracted from the OSM
data. 2) Lines that represent different types of roads, walking areas, sidewalks,
zebra crossings. Each line was defined by its name (for the streets), or its direction
compared to North for differentiating road sidewalks. 3) Points that represent
Landmarks and POIs.
The platform presented two distinct modes: firstly a Control mode that is used
by VE developers, researchers, and Orientation and Mobility instructors. The
Control mode allows developers to create and modify VEs. A key feature of the
Control mode is the program’s ability to import an XML file from Open Street
Map to create a new 3D virtual map and to manually or automatically (via the
222 Kammoun et al.
route selection algorithm from Kammoun et al., 2010) select a path between two
points. This makes it possible to easily import maps of different campuses or cities.
The control mode includes the feedback editor (tactile and auditory feedbacks).
The second Evaluation mode allows researchers and Orientation and Mobility
instructors to record and review the user behaviour in the VE. During an
experimental session, the system recorded in a text file the avatar’s (user) position,
orientation and speed within the VE, along with any interaction between the user
and the system.
For the experimenter display, different textures were applied according to the
type of surface (building, walls, etc.) or the type of line (e.g. tar texture is chosen
for roads, and zebra texture for pedestrian crossing). Figure 22.1a shows an
example of the University of Toulouse environment including the global map in
the lower right hand corner. Figure 22.1b shows a close-up view of the global map
with the actual position of the user (red spot) in the virtual space. The platform was
implemented in C++ code running under Windows 7. It uses an open-source
graphic rendering engine (OGRE 3D), and a cross-platform 3D audio API
(OpenAL) appropriate for displaying 3D spatialised sounds via the headphones. A
collision detection algorithm based on ray tracing was integrated into the
framework so as to allow real time collision detection in complex environments.
a) b)
Figure 22.1. a) A screenshot from the graphic interface of the virtual environment. The
global map is on the lower right hand corner. One can observe different buildings, walking
areas and car park areas. b) Close-up view of the global map, the red point represents the
position of the traveller in the virtual space. Each type of walking area (sidewalk, pedestrian
crossing, car park area, and road) is represented by a distinct colour.
22.4 Navigation and Guidance Process

Before beginning guidance, whether in real or virtual conditions, route selection is
necessary and is usually included in the GIS component. It is defined as the
procedure of choosing an optimal pathway between origin and destination. When
considering pedestrian navigation, the shortest path might be appropriate but
should rely on a GIS database including essential information for pedestrian
mobility (e.g. sidewalks and pedestrian crossings). In addition, brainstorming
sessions revealed that excluding difficult points (e.g. complicated crossroads) as
well as including non-visual landmarks on the itinerary is important when route
selection is made for the Visually Impaired. Using the database described in
chapter 22.3.1 we were able to select an adapted path for a VI pedestrian by using
the algorithm indicated in (Kammoun et al., 2010). When a route was selected,
several sections defined by two successive Itinerary Points (IPs) were generated.
Each section contained a list of Landmarks and POIs extracted from the GIS.
As mentioned in the previous section, virtual 3D sounds, TTS and Spatialised
TTS were used to display instructions and environmental information (see Figure
22.2). Based on simple interaction properties attached to the different scene
elements it will be possible to design and evaluate many guidance processes. A
first instance of direction instruction process was designed by displaying a
spatialised sound on the next IP position according to the traveller position and
orientation.
a) b)
Figure 22.2. a) The green point represents an intermediary itinerary point (IP) (a virtual
spatialised sound is placed at this location). When the point is reached, feedback is given to
the subject and the next IP is activated. This guidance process allowed the user to reach
his/her destination by small, easily reachable steps. b) Blue points represent landmarks or
POIs selected on this path, and extracted from the GIS.
Figure 22.3. a) Representation of a recorded journey around two virtual buildings in the
University campus during an evaluation session with a VI user. Green, blue and red points
represent, respectively, IPs, landmarks, and POIs that were displayed. The dark line
represents the trajectory of the avatar. b) During the evaluation session, navigation speed
and orientation of the avatar were also recorded; each arrow represents speed and orientation
during navigation within the VE.
Landmarks and POIs in the vicinity of the path, which provide the user with
spatial indications about his surroundings were displayed. In this case, a circular
activation field (with an adjustable diameter) was attached to each type of element
224 Kammoun et al.
to trigger an information display. Figure 22.3 shows the representation of recorded

displacements during two preliminary evaluation sessions.
22.5 Discussion and Future Work

This paper presents the design and implementation of a virtual environment
framework in order to evaluate tracking and guidance strategies before their
implementation in an EOA. Based on a geographical database including specific -
and essential - information for pedestrian mobility as well as important Landmarks
and POIs for the VI users, we built a virtual environment in which several user
tracking and guidance strategies will be evaluated safely and systematically in the
laboratory. The most promising strategies will then be tested in the real world with
a prototype of an EOA sharing a large set of components with the VE (GIS
database; route selection, tracking and guidance software; TTS and binaural
synthesis output, see Katz et al., 2010).
22.5.1 User Tracking and Guidance Process

Tracking user location is essential for an efficient guidance process in visually
impaired navigation. For EOAs, geolocation based on GPS is the most common
technique. However, positioning precision with GPS alone is rarely better than 20
metres in many environments. This is particularly the case in cities where urban
canyons between buildings prevent direct line of sight with the satellites. Several
approaches have been developed to improve the GPS precision in such
environments. Dead-reckoning algorithms combine GPS with inertial sensors to
improve estimated user positioning. Differential GPS (see e.g. Loomis et al., 1994)
can provide an accurate positioning, and hence better guidance, but rely on a
network of stations that is not available everywhere. Electronic location identifiers,
such as RFID tags, WLAN networks or Bluetooth beacons have also been used in
indoor environments such as museums, hospitals, and so on. However, they entail a
large scale infrastructure with a specific deployment phase and important
maintenance costs, which is very rarely available in cities. In this context, the use
of a platform simulating real environment and positioning (with added inaccuracy)
seems a promising approach when designing an EOA. Indeed, in this experimental
platform, it is very easy to add in systematic and random noise to the positioning of
the avatar. The resulting positioning will mimic what is observed with real GPS in
different environments. The platform will then be used for testing the robustness
and usability of the different tracking and guidance processes implemented.
22.5.2 Sound Localisation and Haptic Feedbacks

In the platform that we describe, guidance instructions and environmental
description are displayed using a combination of TTS, binaural synthesis and
haptic feedback. There is a well-known issue related to the use of binaural sounds:
the spatialisation must be defined in the head-centred reference frame of the user,
not in that of the external world. However (Katz and Picinali, 2011) showed that
head-tracking was not necessary for 3D sound-based navigation in VE; 3D sound
synthesis based on the movements of the joystick (and hence on the displacement
of the avatar) was sufficient for users to succeed in non-visual navigation tasks.
A rich haptic interaction is provided through the use of a force feedback
joystick device. Many scene elements can generate vibration and force feedback
when they are encountered. This haptic interaction was designed to reproduce the
information that a visually impaired person gathers - including ground texture, step
and obstacle detection - when moving with a white cane or a dog.
22.5.3 Spatial Cognition in the Visually Impaired

Virtual environments appeared to be effective in the transfer of cognitive mapping
to real environments, even for visually impaired users (see e.g. Lahav and
Mioduser, 2003). Our aim was different in that the platform is designed for
evaluating robustness of tracking and improved guidance for the visually impaired
(including map adaptation, route selection, direction instruction and space
description). Indeed, as we previously stated, this platform will ease the tests of
route selection - comparing for instance the automatically selected route with the
users’ preference - or robustness of tracking algorithms by adding systematic
and/or random noise to the position of the avatar. But, importantly, we will also
modify the quality and the quantity of direction instructions, landmarks and POIs
mentioned during the journey. We will use different experimental designs to test
the efficiency of the guidance itself (number of errors and time to succeed), but
also the mental maps acquired during the journey. These maps will be compared
with maps acquired during mobility in real environments, as well as with
topographic maps (see Lahav and Mioduser, 2008, for indoor real vs. virtual
exploration). Our aim is to finally select the strategies that enhance not only the
user-centred representations used in turn-by-turn instructions but also the global
spatial knowledge of the explored environment.
22.6 References
Brock A, Vinot J-L, Oriola B, Kammoun S, Truillet P, Jouffrais C (2010) Méthodes et outils
de conception participative avec des utilisateurs non-voyants. In: Proceedings of the 22nd
Francophone Conference on Human-Computer Interaction, Luxembourg
Gaunet F (2006) Verbal guidance rules for a localised wayfinding aid intended for blind-
pedestrians in urban areas. Universal Access in the Information Society, 4(4): 338-353
Golledge RG, Klatzky RL, Loomis JM, Speigle JM, Tietz J (1998) A geographical
information system for a GPS based personal guidance system. International Journal of
Geographical Information Science, 12(7): 727-749
226 Kammoun et al.
Kammoun S, Macé MJ-M, Oriola B, Jouffrais C (2011) Toward a better guidance in

wearable electronic orientation aids. In: Proceedings of the 13th IFIP TC13 International
Conference on Human-Computer Interaction, Lisbon, Portugal
Kammoun S, Dramas F, Oriola B, Jouffrais C (2010) Route selection algorithm for blind
pedestrian. In: Proceedings of the International Conference on Control, Automation and
Systems (ICCAS 2010), Gyeonggi-do, Korea
Katz BFG, Picinali L (2011) Spatial audio applied to research with the blind. In: Strumillo P
(ed.) Advances in sound localization. Available at: http://www.intechopen.com/articles/
show/title/spatial-audio-applied-to-research-with-the-blind (Accessed 9 November 2011)
Katz BFG, Truillet P, Thorpe S, Jouffrais C (2010) NAVIG: Navigation assisted by artificial
vision and GNSS. In: Proceedings of the Workshop on Multimodal Location Based
Techniques for Extreme Navigation (Pervasive 2010), Helsinki, Finland
Lahav O, Mioduser D (2003) A blind person’s cognitive mapping of new spaces using a
haptic virtual environment. Journal of Research in Special Educational Needs, 3(3): 172-
177
Lahav O, Mioduser D (2008) Haptic-feedback support for cognitive mapping of unknown
spaces by people who are blind. International Journal of Human-Computer Studies 66(1):
23-35
Loomis JM, Golledge RG, Klatzky RL, Speigle J (1994) Personal guidance system for the
visually impaired. In: Proceedings of the ACM Conference on Assistive Technologies
(ASSETS ’94), Marina Del Rey, CA, US
Loomis JM, Golledge RG, Klatzky RL (1998) Navigation system for the blind: Auditory
display modes and guidance. Presence: Teleoperators and Virtual Environments, 7(2):
193-203.
Loomis JM, Klatzky RL, Golledge RG (2001) Navigating without vision: Basic and applied
research. Optometry and Vision Science, 78(5): 282-289
Mereu SW, Kazman R (1996) Audio enhanced 3D interfaces for visually impaired users. In:
Proceedings of the ACM International Conference on Human Factors in Computing
Systems (CHI 1996), Vancouver, Canada
Middlebrooks JC, Green DM (1991) Sound localization by human listeners. Annual Review
of Psychology, 42: 135-159
Pielot M, Henze N, Heuten W, Boll S (2008) Evaluation of continuous direction encoding
with tactile belts. In: Proceedings of the International workshop on Haptic and Audio
Interaction Design, Jyväskylä, Finland
Ross DA, Blasch BB (2000) Wearable interfaces for orientation and wayfinding. In:
Proceedings of the International conference on Computers and accessibility, Arlington,
VA, US
Sanchez J, Hassler T (2006) AudioMUD: A multi-user virtual environment for blind people.
In: Proceedings of the International Workshop on Virtual Rehabilitation, Manhattan, NY,
US
Sánchez J, Tadres A (2010) Audio and haptic based virtual environments for orientation and
mobility in people who are blind. In: Proceedings of the International conference on
Computers and accessibility, Orlando, US
Sanchez J, Tadres A, Pascual-Leone A, Merabet L (2009) Blind children navigation through
gaming and associated brain plasticity. In: Proceedings of the International Conference on
Virtual Rehabilitation, Haifa, Israel
Schloerb DW, Lahav O, Desloge JG, Srinivasan MA (2010) BlindAid: Virtual environment
system for self-reliant trip planning and orientation and mobility training. Haptics
Symposium, Massachusetts, MA, US
Chapter 23
Spatial Clues for Orientation: Architectural

Design Meets People with Dementia
I. Van Steenwinkel, C Van Audenhove and
A. Heylighen
23.1 Introduction
The physical environment holds great potential to improve the well-being of
people with dementia (Calkins et al., 2001; Van Audenhove et al., 2003;
Sternberg, 2009; Van Audenhove et al., 2009). However, when designing
environments for them, architects are faced with a lack of adequate design
knowledge. On the hypothesis that the perspectives of people with dementia have
the potential to expand architects’ design expertise (Zeisel, 2001), our research
aims to gain a better understanding of their spatial experiences in order to achieve
this end. This paper outlines how this overall objective has become more
articulated into more specific research questions through preliminary research:
• by reviewing literature in the fields of anthropology, psychology,
phenomenology, human geography, architectural theory, supplemented
with a study of (auto)biographies of people with dementia;
• by visiting 22 residential and care environments for people with dementia
in Flanders and one in the Netherlands;
• by talking to people with dementia, their family and professional care
givers and other professionals;
• by conducting open interviews with five architects, experienced in
designing residential and care environments for people with dementia,
about the strategies they used in former design projects;
• by conducting voluntary work, both in a day care centre and in a residential
care centre for people with dementia, one day a week during two months,
to become familiar with their daily life.
Secondly this paper explains which research methods are chosen to address the
proposed research questions, and why.

228 Van Steenwinkel et al.
23.2 Context
Dementia is a syndrome of progressive memory impairment and loss of other
cognitive functions (American Psychiatric Association, 2000). This causes a very
particular way of experiencing and negotiating space (Godderis, 1992). Family,
friends and professional care givers often face a lack of insights into the way
people with dementia perceive the environment. It can be difficult, for example, to
find out why someone unexpectedly - and seemingly without provocation - gets
anxious, suspicious, or aggressive, how a person can get lost in a familiar
environment (Friel McGowin, 1993), or why people with dementia sometimes just
sit and stare for a long period of time seemingly detached from what happens
around them (Boden, 1998).
At the same time, we expect that there is much to learn from people with
dementia. As the main thread of the research conducted by our team, we consider
people with different abilities and/or conditions - in this case people with dementia
- as experts in perception and use of the (built) environment. They can bring
forward the very different ways in which people in a diverse and aging society
experience and negotiate their living environment and, thus, they can expand and
refine architects’ knowledge on how to design that environment.
In case of designing a living environment for people with dementia, the client
often differs from the future user, i.e. people with dementia. If concerned with this
future user, the client tries to communicate the user’s needs and desires to the
architect(s). Although clients often have several years of experience in caring for
people with dementia and the (built) environments they live in, interviews with
architects reveal that clients can articulate only few insights in the spatial needs of
people with dementia. For example, clients tend to emphasise the importance of
“normality” and “homeliness”, characteristics which architects find too vague to
work with. On the other hand, here lies an important role for architects, since we
expect that their spatial knowledge can afford new insights in the daily experiences
of people with dementia. In practice, however, architects often do not have or take
the time to explore this in depth.
Following from the above, we set out to explore the spatial experiences of
people with dementia from an architectural point of view. Moreover, we assume
that from this exploration we may also learn about the spatial experiences of
people in general. Indeed, people with dementia are persons in the first place, who
undergo a relative slow dementing process. Their often unrestrained and
spontaneous behaviour (Zeisel, 2001) may reveal how other people secretly or
unconsciously experience a certain situation.
23.3 Research Focus

23.3.1 Orientation
Each person with dementia experiences a particular situation in a unique way.
There are several types of dementia and different, but not clearly discriminated,
Spatial Clues for Orientation: Architectural Design Meets People with Dementia 229
stages in the dementing process, which may succeed each other at different rates.
The disease can strike people of different ages and all kinds of backgrounds
(Godderis, 1992). Since we are faced with such a great diversity of people who
often live together in a group housing facility, we do not intend to focus on a
delimited target group, e.g. women with early-onset Alzheimer’s. Instead we focus
on one important aspect which all people with dementia have in common, i.e.
disorientation in time, space and identity, due to memory loss (Godderis, 1992).
People with dementia, may find their sense of time becomes upset or even lost.
Now and then they do not know which (time of the) day or year it is, as Christine
Boden, a women with early-onset Alzheimer’s, witnessed:
“I don’t seem to have space in my brain for that sense of ‘Thursday-ness’ (or
whatever day it might happen to be), or ‘April-ness’ or ‘1981-ness’.”
(Boden, 1998)
One of the possible consequences is that their (daily) routines may get mixed
up. A person with dementia may, for example, want to go shopping in the middle
of the night (Braam, 2005). Something similar may happen concerning orientation
in space. A person with dementia may get lost in space and be baffled when, in the
morning, “everything seems new” (Braam, 2005). Additionally, people with
dementia show a particular way of knowing people, as Christine Bryden described:
“You see, I did not know their name, whether they were married or not, whether
they had children, if they had a job. I knew nothing about them, nothing in the
‘normal’ sense of how you know people and recognise them. The way I know people
is in a spiritual and emotional way. There’s a knowing of who a person really is
right at their core. But I have no idea who they are, in terms of who they are meant
to be in your world, of cognition and action, and labels and achievement.”
(Bryden, 2005)
The fact that people with dementia sometimes are disoriented in time as well in
space and identity may not be a surprise, since - according to several authors -
time, space and identity are interrelated dimensions. Edward Hall (1969) touched
upon the idea that the differentiation of time and space as two distinct dimensions
is only an arbitrary one, since in lived experiences they are actually “inextricably
bound up in each other” (Hall, 1969). Christian Norberg-Schulz (1971) also
explained that “perception mediates a world which could also very well be
described as ‘events in a four-dimensional space-time’.”
Moreover, Hall (1969) points out that space, and thus time, is also related to
identity:
“Man’s sense of space is closely related to his sense of self, which is in an intimate
transaction with his environment. Man can be viewed as having visual, kinesthetic,
tactile, and thermal aspects of his self which may be either inhibited or encouraged
to develop by his environment. “
(Hall, 1969)
Because of these interrelations we henceforth use the term ‘orientation in time-
space-identity’. Thus, we do not consider orientation merely as a means for way-
finding, i.e. knowing how to go from one geographical location to another. We use
‘orientation’ in a broad sense of the word. Including identity as one dimension of

orientation, we also consider questions like: Can I be myself here? Does this place
detach me from who I am, my past and my hoped-for future? Can I find a place of
my own? Do I feel at home in this place? In fact, ‘home’ is a special reference
point in all dimensions of orientation, i.e. time, space and identity. Indeed, one’s
home is a particular meaningful place (Norberg-Schulz, 1971; Tuan, 1977;
Madanipour, 2003) and it has an important role in colouring one’s identity (Tuan,
1977; Madanipour, 2003; Chapman, 2006). Yi-Fu Tuan (1977) elicits the time
dimension of home by writing that “in an ideal sense home lies at the center of
one’s life, and center (we have seen) connotes origin and beginning.”
Exploring how an entity like home carries a sense of time, space and identity
may reveal how, for a particular person, it can serve as a point of reference for
orientation. The particular experiences of people with dementia may elicit new
insights, or vice versa, such an exploration may reveal why people with dementia
sometimes can (not) orient themselves. After all, people with dementia not seldom
express the desire to go home, even when in fact they are at the location where
they currently live.
Based on these considerations, we delineate our research question a little more
precisely: How can the physical environment allow or impede a person to orient
their self in time-space-identity? In other words: How can physical entities provide
a sense of time, space and identity, through which they can be points of reference
for orientation? Hence, we look at how one dimension of orientation, i.e. the
physical entities of space, contributes to or hinders the overall orientation, i.e.
orientation in time-space-identity.
23.3.2 Movement
If we take our research question a step further, the question arises: How does a
person develop a sense of time, space and identity? Literature study revealed that
‘movement’ is a very important, if not an essential key to the development of a
sense of time, space, and identity. We found strong indications that a focus on
movements may yield an understanding of the lived experiences of people with
dementia.
Tim Ingold (2000) contends that “movement is the very essence of perception”
and Tuan (1977) explains the role of movement in orientation in time and space
(and therefore also in identity) by writing:
“We can have sense of space because we can move and of time because, as
biological beings, we undergo recurrent phases of tension and ease. The movement
that gives us a sense of space itself is the resolution of tension. When we stretch our
limbs we experience space and time simultaneously - space as the sphere of freedom
from physical constraint and time as duration in which tension is followed by ease.”
(Tuan, 1977)
Movements can be of different kinds and of different scales. For example, the
possible bodily movements in a room influence how a person experiences that
room. At a bigger scale, the daily commuting between home and office has a part
in the ‘image’ a person develops of their living environment. Cycles like night and
day or the seasons are also movements, which “are embodied, incorporated into
our very constitution as biological organisms” (Ingold, 2000). Such movements
contribute to orientation. Being out of tune with them may be stressful and/or
confusing, for example, when a person with dementia gets up in the middle of the
night and thinks he/she needs to buy some food, but finds all stores closed (Braam,
2005), or when, in the late afternoon, a person feels it is time to go home, but is
hindered by carers or locked doors. Movements, sequences of tension and ease,
should not only be considered as physical, but also cognitive and emotional, e.g.
being ‘moved’ by a(n emotionally charged) conversation with friends or family
versus a peaceful time on your own.
Piaget and Inhelder (1971) describe that a comprehension of the physical
environment evolves from internalised actions rather than from mental images. We
know things not only by using our brain, but also by incorporating the movements
of other parts of our body. The following quote from a person with dementia
illustrates that his hands ‘know’ his PIN code when standing in front of the
keyboard, while his head could not remember it in advance:
“En route, I try to remember the PIN code. […] When I stand in front of the desk, I
know.”
(Braam, 2005)
The development of a sense of time also requires movements, and more
precisely the co-ordination of moving entities (like your own body, that of others,
or objects) at different velocities (Piaget, 1969).
Norberg-Schulz (1971) - drawing on the work of Otto Friedrich Bollnow
(1963) - describes the importance of movement to human beings as follows:
“How we get from one place to another is a basic aspect of man’s being in the
world. […] Life itself can be understood as movement from one condition to
another. This movement is incessant and continuous, but it has rhythm and form.
[…] Furthermore, man is part of a system of natural rhythms, such as night and day,
the change of seasons, and his own ‘ages’.
(Norberg-Schulz, 1971)
Therefore, movement seems to be a basic aspect of a person’s well-being or ill-
being. In the context of care for people with dementia we thus consider it valuable
to investigate their ‘movements’, through time, space and identity and the way
their built environment may reveal or hide cycles like night and day or possible
other events that mark particular points in time.
Consequently, we reformulate our research question as follows: How can the
physical environment as a motor medium afford or impede a person’s orientation
in time-space-identity? Setting out our exploration by considering the experiences
of people with dementia raises several sub-questions: Which activities, pauses,
events make up or are left out of the ‘timescape’ of a person with dementia,
supporting or hindering orientation in time? Which physical features induce
movements that constitute ‘the image’ of the environment for people with dementia
and, therefore, operate as clues for orientation in space? How can the physical
environment afford a place of one’s own affording orientation in one’s own
identity? And how can the physical environment afford the feeling of coming or
being home, supporting orientation in time-space-identity? By analogy with
‘landscape’. Ingold (2000) uses the term ‘taskscape’, yet we prefer ‘timescape’
because it refers more explicitly to the dimension of time. By analogy with ‘the
image of the city’, described by Kevin Lynch (1960), though we will not limit our
study to only visual aspects of space.
23.4 Methodological Issues

Addressing the proposed research questions requires consideration of both the
physical environment and people living with dementia and these two in relation to
each other. We plan to tackle the research questions in two ways: firstly, through
ethnographic research that is a combination of an analysis of the physical
environment, interviews and observations, and secondly, through ‘research by
design’.
23.4.1 Ethnographical Research

Studies on people with dementia typically focus on one particular housing context,
e.g. either large scale residential settings (nursing homes or Alzheimer‘s special
care units) (Sloane et al., 1998), or small group living facilities ( Elmståhl et al.,
1997; Van Audenhove et al., 2003), which limits the transferability of the findings
to other contexts. By contrast, our research project aims to study the spatial
experience of people with dementia across different contexts and housing schemes.
Possibly, these differences can even accentuate how people with dementia can
(not) orient themselves in time-space-identity. We take into account persons at
their homes - like individual houses, apartments, service flats - and persons in
group housing facilities - like small group living facilities or traditional large-scale
nursing homes. These settings are analysed in terms of five spatial aspects:
• spatial articulation, which deals with social aspects of space, like being
(semi-)private or public;
• enclosure, which deals with how spaces are delineated by physical entities
like walls, windows, level changes, furniture, signs, etc.;
• sensory qualities, like colour, sound, smell, warmth, texture, etc.;
• materials, which will be analysed not only for their sensory qualities but
also for the meanings they may evoke;
• form, measurements and proportions of rooms, openings in rooms, outdoor
places, objects, etc.
The list is based on spatial aspects described by Ola Nylander (1999), which
have been modified in a study on living environments for older people (without
dementia). We do not consider this list of spatial aspects fixed, but allow it to be
modified during the research if necessary.
The spatial analysis is complemented with both interviews and observations.

Through interviews we try to explore how entities in a person’s physical
environment carry a sense of time, space and identity. All too often, cognitive
impairment is assumed to obliterate the ability of people to evaluate and share their
evaluations with others (Cotrell and Schulz, 1993). Interviewing people with
dementia may be challenging, but it is not impossible. In fact it has been done
many times before. Louise Nygård (2006) draws on more than 10 years of
ethnographic research to reflect on and make suggestions for how the scientific
exploration of the experiences of people with dementia may be undertaken. She
suggests that “a combination of qualitative observations and adapted interviews
may make it possible for people with dementia to participate as research
informants” (Nygård, 2006).
Interviews should be adapted to the participant’s capacities, in terms of
formulating questions and interview duration (Nygård, 2006). Therefore, the
researcher must be very flexible. Number, length, focus and context of interviews
may be different with every participant (Nygård, 2006). In our research, the
number of interview participants will be decided based on a pilot study.
Observations will offer extra material on the daily life and experiences of
people with dementia by mapping out (i.e. making notes and drawings) the
‘movements’ by persons with dementia and their environment. These ‘maps’ are
made over a period of time and include accounts of what people do and say, when
they use a space, where they sit, the way they walk, etc.
23.4.2 Research by Design

In a second part, we will confront the perspective and appreciation of people with
dementia related to the physical environment with the perspective of architects, and
in particular with their designerly ways of knowing (Cross, 2006). The approach
used in this work package shows similarities with research by design in the sense
that the activity of designing is used as a way to generate knowledge. However,
unlike in typical instances of research by design (Pedgley, 2007), the researcher
will not perform this design activity herself, but will call in others’ design
activities. We plan to involve architects in the research process in two ways:
• by challenging architects to freely (i.e. without any precondition)
reconsider (the design of) space from the perspective of people with
dementia, introduced through information from the ethnographic fieldwork;
• by introducing insights from the ethnographic fieldwork as input for a real-
world design assignment.
In this way we call in architects’ expertise to tackle our research questions. The
thoughts of these architects are expected not only to offer a glimpse of future
possibilities, were architectural design to meet people with dementia, but also to
further our understanding of the spatial experience of people with (and without)
dementia. Unlike scientists who set out specifically to study a problem/issue,
architects (like other designers) learn about the nature of the problem/issue largely
as a result of trying out solutions (Lawson, 1998; Cross, 2006).
Moreover, the design outcomes of both approaches are presented to people with
dementia, their relatives and/or care givers. Indeed, former research projects have
shown that the evaluation of (unrealised) designs by possible users can elicit
additional insights in their experiences, like in research projects dealing with the
experiences of visually impaired people (Vermeersch et al., 2011).
23.5 Conclusions
The general aim of our research is to better understand the spatial experiences of
people with dementia, firstly because the physical environment holds great
potential to improve their well-being, and secondly because of the assumption that
their perspective could expand architects’ design expertise.
Based on preliminary research we figured out that - beside space - dimensions
of time and identity should be included in the research, since time, space and
identity are inextricably bound up with each other. Moreover, people with
dementia, because of their memory impairments, find difficulties in orientating in
all three dimensions. A literature study elicited useful concepts and theories to
explore the experiences of people with dementia. Most importantly, many authors
consider movement essential to perception. That is why we focus on how the
physical environment as a motor medium assists or impedes people in orientating
themselves in time-space-identity.
Our research covers several (if not all) types of dementia, several types of
housing facilities, and people with different backgrounds. In this way, insights can
be transferred to different contexts. Nevertheless, we are aware that other
important factors, e.g. socio-cultural background, have a part in how people
negotiate space. For that reason material from the fieldwork should be interpreted
cautiously with regard to the transferability of insights to different contexts.
We hope to add new insights on important aspects of how people with frailty
and cognitive impairment negotiate space and that our results will help designers to
improve the orientation and wellbeing of people with dementia.
The research reported here received support from the European Research Council
under the European Community’s Seventh Framework Programme (FP7/2007-
2013)/ERC grant agreement No. 201673. The authors would like to thank the
persons working in the residential care centres with whom we talked and who
showed us around in the buildings where they work, the architects who participated
in interviews, and especially the persons with dementia, their friends, relatives and
care givers for sharing time and experiences.
23.7 References
American Psychiatric Association (2000) Diagnostic and statistical manual of mental
disorders: DSM-IV-TR. American Psychiatric Publications, Washington, DC, US
Van Audenhove C, Declercq A, De Coster I, Spruytte N, Molenberghs C, Van den Heuvel B
(2003) Kleinschalig genormaliseerd wonen voor personen met dementie. Garant,
Apeldoorn, The Netherlands
Van Audenhove C, Spruytte N, Detroyer E, De Coster I. Declercq A, Ylieff M et al. (2009)
De zorg voor personen met dementie: Perspectieven en uitdagingen. Koning
Boudewijnstichting, Brussels, Belgium
Boden C (1998) Who will I be when I die? HarperCollins, Melbourne, Australia
Bollnow OF (1963) Mensch und raum. Kohlhammer, Stuttgart, Germany
Braam S (2005) Ik heb Alzheimer: Het verhaal van mijn vader. Nijgh & Van Ditmar,
Amsterdam, The Netherlands
Bryden C (2005) Dancing with dementia: My story of living positively with dementia.
Jessica Kingsley, London, UK
Calkins M, Sanford JA, Proffitt MA (2001) Design for dementia: Challenges and lessons for
universal design. In: Preiser WFE, Ostroff E (eds.) Universal design handbook. McGraw
Hill Professional, NY, US
Chapman S (2006) A “new materialist” lens on aging well: Special things in later life.
Journal of Aging Studies, 20(3): 207-216
Cotrell V, Schulz R (1993) The perspective of the patient with Alzheimer’s disease: A
neglected dimension of dementia research. The Gerontologist, 33(2): 205-211
Cross N (2006) Designerly ways of knowing. Birkhäuser Verlag, Basel, Switzerland
Elmståhl S, Annerstedt L, Ahlund O (1997) How should a group living unit for demented
elderly be designed to decrease psychiatric symptoms? Alzheimer’s disease and
Associated Disorders, 11(1): 47-52
Friel McGowin D (1993) Living in the labyrinth: A personal journey through the maze of
Alzheimer’s. Thorndike Press, Thorndike, ME, US
Godderis J (1992) Handboek geriatrische psychiatrie. Garant, Leuven. The Netherlands
Hall E (1969) The hidden dimension. Doubleday, NY, US
Ingold T (2000) The perception of the environment: essays on livelihood, dwelling and skill,
Routledge, London, UK
Lawson B (1998) How designers think, 3rd edn. Butterworth Architecture, Oxford, UK
Lynch K (1960) The image of the city. The MIT Press, MA, US
Madanipour A (2003) Public and private spaces of the city, 1st edn. Routledge, London, UK
Norberg-Schulz C (1971) Existence, space and architecture. Studio Vista, London, UK
Nygård L (2006) How can we get access to the experiences of people with dementia?
Scandinavian Journal of Occupational Therapy, 13(2): 101-112
Nylander O (1999) The home as architecture. Bok & Bild, Gothenburgh, Sweden
Pedgley O (2007) Capturing and analysing own design activity. Design Studies, 28(5): 463-
483
Piaget J (1969) The child’s conception of time. Routledge and Kegan Paul, London, UK
Piaget J, Inhelder B (1971) The child’s conception of space, 4th edn. Routledge and Kegan
Paul, London, UK
Sloane PD, Mitchell CM, Preisser JS, Phillips C, Commander C, Burker E (1998)
Environmental correlates of resident agitation in Alzheimer’s disease special care units.
Journal of the American Geriatrics Society, 46(7): 862-869
Sternberg E (2009) Healing spaces: The science of place and well-being. The Belknap Press
of Harvard University Press, Cambridge, MA, US
Tuan Y-F (1977) Space and place: the perspective of experience. Edward Arnold
(Publishers) Ltd, London, UK
Vermeersch P-W, Nijs G, Heylighen A (2011) Mediating artifacts in architectural design: A
non-visual exploration. In: Leclercq P, Heylighen A, Martin G (eds.) Designing together.
CAADfutures. Les Editions de l’Université de Liège, Liège, Belgium
Zeisel J (2001) Universal design to support the brain and its development. In: Preiser WFE,
Ostroff E (eds.) Universal design handbook. McGraw Hill Professional, NY, US
Index of Contributors
e
Afacan Y .....................85 Cirba N ...................... 43 Goldhaber T.S .......... 105

Akeley C ...................185 Clarkson P.J ....... 55, 105, Goodman-Deane J....... 55
.................. 115, 133, 145
Almuhim M ..............185 Gorman M ................ 185
Cobb S ..................... 155
Annemans M ................3 Gregory C ................. 185
Conduit G ................... 93
Attanayake D ............125
Cooke T ..................... 23
Henderson H .............. 75
Corso A ...................... 93
Barmoy P ..................185 Herriott R ................. 165
Crouch D .................. 185
Beavan P ...................185 Heylighen A ........ 3, 207,
.................................. 227
Beck C ......................185
Dalke H....................... 93 Holt R.......................... 33
Bhakta B .....................33
Dehmer G ................ 185 Horberry T ................. 23
Biswas P ...................195
Denholm-Price J ....... 125 Hosking I .................... 55
Blair A ......................185
Dong H ...................... 13 Hunter G ................... 125
Bortz A .....................185
Bracewell R.H ..........133
Edlin-White R........... 155 Jokisuu E .................. 115
Bradley B ..................185
Jordan P.W ................. 13
Bradley M....................55 Floyde A .................. 155
Jouffrais C ................ 217
Carter M ...................185 Gallagher J.F .............. 33

Ceccacci S ..................65 Kammoun S ............. 217
Germani M ................. 65
Chakraborty S .............75 Goddard N ............... 175 Kharal N ..................... 43

DOI: 10.1007/978-1-4471-2867-0, © Springer-Verlag London 2012
238 Index of Contributors
Langdon P.M ......55, 105, Oriola B ................... 217 Van Audenhove C........ 3,
...........115, 133, 145, 195 .................................. 227
Lanier E ....................185 Van Steenwinkel I .... 227
Patmore J.J ............... 133
Lazar J .................75, 185 Vermolen H................... 3
Pfluegel E ................. 125
Levesley M.C .............33
Preston N ................... 33
Lewthwaite S ............155 Waller S.D................... 55
Wang J ..................... 155
Riaz A ........................ 93
Macé M.J-M .............217 Weightman A.P.H ...... 33
Riedel J .................... 155
McIntee A .................185 Weir R ........................ 75
Ritgert D .................. 185
Mengoni M .................65 Wells J ...................... 185
Rogers Jr. R ............. 185
Mieczakowski A .......133 Wentz B ............. 43, 185
Rosenwald S ............ 185
Mon-Williams M ........33 Willis C .................... 185
Moran J .......................43 Sizemore B ................ 75 Wingo-Jones K ......... 185
Slate M ....................... 43
Nelson Jr. R ..............185 Sullivan S ................. 185 Yatto T ..................... 185
Nickpour F...................13
Nicolle C ..................175 Tenneti R .................... 55 Zitkus E .................... 145

978 1 4471 2867 0

Uploaded by

Copyright:

Available Formats

978 1 4471 2867 0

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

978 1 4471 2867 0

Uploaded by

Copyright:

Available Formats

Designing Inclusive Systems

Patrick Langdon John Clarkson

Peter Robinson Jonathan Lazar

Designing Inclusive Systems

ISBN 978-1-4471-2866-3 e-ISBN 978-1-4471-2867-0

British Library Cataloguing in Publication Data

Library of Congress Control Number: 2012933436

Ó Springer-Verlag London 2012

Cover design: eStudio Calamar S.L.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

The Cambridge Workshops on Universal Access and Assistive Technology

proving ground is required in real-world application and industry. This proving

There are five main themes:

V. Designing Inclusive Architecture highlights specific cases, such as

Pat Langdon, John Clarkson, Peter Robinson

The CWUAAT Editorial Committee

List of Contributors …………………………………………………………..xiii

Part I Designing for the Real-world

2. Inclusive Bus Travel - A Psychosocial Approach

3. Safe and Inclusive Design of Equipment Used in the

4. Deploying a Two-player System for Arm Rehabilitation

5. Evaluating the Accessibility and Usability of Blogging

Part II Measuring Demand and Capabilities

7. How to Use Virtual and Augmented Reality Techniques to

8. Development and Evaluation of Sonified Weather Maps for

9. Achieving Inclusion in Public Spaces: A Shopping Mall Case

10. Visibility Prediction Software: Five Factors of Contrast

Part III Designing Cognitive Interaction with

12. A Framework for Studying Cognitive Impairment to

13. Interactive Error Correction Using Statistical Language Models

14. Understandable by Design: How Can Products be Designed

Part IV Design for Inclusion

16. From Guinea Pigs to Design Partners: Working with Older

17. When Users Cannot be Included in Inclusive Design

18. What is Good Design in the Eyes of Older Users?

19. Equal Access to Information? Evaluating the Accessibility of

20. Clustering User Data for User Modelling in the GUIDE

Part V Designing Inclusive Architecture

22. Designing a Virtual Environment Framework for Improving

23. Spatial Clues for Orientation: Architectural Design Meets

Index of Contributors …………………………………………………..…..237

Afacan Y., Department of Interior Architecture and Environmental Design,

Bradley B., Department of Computer and Information Sciences, Universal

Gorman C., Department of Computer and Information Sciences, Universal

Nickpour F., Inclusive Design Research Group, School of Engineering and

Willis C., Department of Computer and Information Sciences, Universal Usability

Designing for the Real-world

Hospital Reality from a Lying Perspective:

1.1 Real Buildings, Real Experiences

P. Langdon et al. (eds.), Designing Inclusive Systems, 3

1.2 Sensory Reality

1.3 Insight/Inside Spatial Experience

1.4 “Talking” Patients, “Speaking” Hospitals

1.4.2 How Patients “Talk”

1.4.3 What Hospitals “Say”

the experience of the building without being necessarily related to visual

Suggesting alterations to the hospital layout is not restricted to drawing plans.

1.5 Conclusions and Future Work