Papers Crowdsourcing
Papers Crowdsourcing
Papers Crowdsourcing
By
W
IE
A dissertation submitted in partial fulfillment of
the requirements for the degree of
EV
DOCTOR OF PHILOSOPHY
PR
MAY 2016
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
W
IE
EV
ProQuest 10139706
Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
W
IE
EV
PR
____________________________
KD Joshi, Ph.D, Chair
____________________________
Arvin Sahaym, Ph.D.
W
IE ____________________________
Kenneth Butterfield, Ph.D.
EV
____________________________
Terence Saldanha, Ph.D.
PR
ii
CROWDSOURCING IT WORK: A THREE-FOLD PERSPECTIVE FROM
Abstract
W
Chair: KD Joshi
IE
This dissertation examines the phenomena of crowdsourcing within the context of the IT
three perspectives: the worker (or labor supply), the buyer of technology services (or labor
demand) and the marketplaces that facility the buyer-seller transaction. It explores how
PR
crowdsourcing can be applied to technology tasks. This dissertation will be structured in a three
essay format. Essay one explores the technology crowdsourcing phenomena from a
“crowdworker” perspective. This essay examines technology crowdwork from a career anchors
perspective, and highlights the potential role of crowdsourcing in expanding the technology
workforce to additional sources of worker capacity. Essay one establishes the theories that
engagements, and utilizes Schein’s Careen Anchors to examine the motivations of workers
technology enabled collaborative work environments. Essay two uses IT Flexibility theory to
iii
examine the relationship between IT flexibility, the rigidity of IT administration and an
organization’s choice to use crowdsourcing as means of IT services delivery. Essay two uses
survey data to show that higher levels of IT flexibility are associated with higher levels of usage
of crowdsourcing. Essay three uses a design science perspective to examine the ability of
crowdsourcing marketplace platforms to meet the needs of IT service buyers and IT service
workers as identified in Essay’s one and two. This essay identifies how simplification of
procurement and management processes could make crowdsourcing more attractive to firms as a
W
IE
EV
PR
iv
Table of Contents
Page
ABSTRACT ................................................................................................................................... iii
INTRODUCTION ...........................................................................................................................1
W
Abstract ........................................................................................................................................4
IE
Introduction ..................................................................................................................................4
EV
Results ........................................................................................................................................20
Discussion ..................................................................................................................................36
Summary ....................................................................................................................................40
Appendix 1 .................................................................................................................................42
Appendix 2 .................................................................................................................................48
v
ESSAY TWO: IT FLEXIBLITY AND THE CHOICE TO USE CROWDSOURCING FOR IT
DELIVERY....................................................................................................................................49
Introduction ................................................................................................................................49
Results ........................................................................................................................................83
W
Discussion ..................................................................................................................................85
IE
Contributions & Limitations ......................................................................................................86
EV
Summary ....................................................................................................................................87
Introduction ................................................................................................................................89
Results ......................................................................................................................................103
Discussion ................................................................................................................................110
Summary ..................................................................................................................................113
vi
List of Tables
Characteristics of Sourcing Environments.......................................................................................9
W
Essay Two Sample Demographics.................................................................................................62
vii
List of Figures
Saturation Graph ............................................................................................................................19
W
Model Results ...............................................................................................................................84
viii
INTRODUCTION
As demands for Information Technology (IT) services continue to escalate companies are
technology related fields, coupled with poor career retention, compounded by pending retirements
in the existing technology workforce support a confluence of forces that will continue to spur
headwinds for IT organizations to meet staffing requirements over the next ten years (Bosworth et
al. 2013).
In the face of these headwinds crowdsourcing is an emerging technique for attracting workers.
W
Crowdsourcing has been defined in a variety of ways. Common themes among crowdsourcing
definitions include utilizing labor from outside the traditional boundaries of the firm to complete
IE
a task (Kaganer et al. 2013). Crowdsourcing has been defined as broadly as virtually any type of
internet based collaboration activity for sourcing work (Estellés-Arolas et al. 2012). In some
EV
instances the contributions of the nontraditional workers are uncompensated, such as contributors
to the Wikipedia website or traditional open source software development projects, and at other
PR
times the contributions of the nontraditional workers are compensated, such as through Amazon’s
Mechanical Turk (Doan et al. 2011). While crowdsourcing has been seen as a means by which
software development capacity can be managed (Kim et al. 2012), successful use of crowdsourcing
in the delivery of software development may require organizations to think differently about
technologies have enabled dramatic change not just in business processes, but also in the
underlying business models that firms leverage to bring value to their stakeholders.
1
This dissertation proposes a research agenda that seeks to take a multifaceted approach to
examining the role of crowdsourcing in expanding the technology workforce. In order for a vibrant
marketplace to exist three elements must be in place: buyers, sellers and a market to facilitate
transactions. This research will examine the technology crowdsourcing phenomenon from each
of these three perspectives using both multi-method and multi-level techniques. Essay one will
examine the crowdsourcing phenomenon from the perspective of the technology worker, or the
seller of IT services. The research will examine the interests and motivations for technology
workers who participate in crowdsourcing platforms. The first essay will seek to address the
following question from the technology crowdworkers’ perspective, Why do technology workers
W
participate in IT crowdsourcing communities? Significant theoretical work has been conducted
IE
describing the variation that can exist in technology career paths (Lee 2002). The first essay
Essay two will explore the attitudes, usage intentions and usage levels of IT crowdsourcing
services by IT services buyers. Technology standardization driven by industry trends such as cloud
PR
computing and mobile devices facilitates the creation of common development environments,
making collaboration between loosely knit crowdworkers possible (Doan et al. 2011). Increasing
levels of technology standardization have been found to facilitate the use of third parties in
software development. Industries with higher levels of standards adoption have been found to
have a greater propensity to engage 3rd party contractors in software development (Sahaym et al.
2007). Architecture governance programs, such as service oriented architecture have been found
to assist in the development of enterprise capabilities across business units (Boh et al. 2007). It is
posited that the same architectural standards that facilitate flexibility and distributed intra-firm
development, can be used to support crowdsourcing engagements. I argue that the degree to which
2
a firm has adopted standardized infrastructures and flexible architectures will shape a firm’s
standardize IT practices, rigidity may develop in vendor management practices that decrease a
firm’s ability to take advantage of emerging sourcing techniques such as crowdsourcing. Building
on this thesis, essay two will examine the following research question: What factors shape an
While the motivations of crowdworkers and the usage intentions of IT services buyers are
important elements to the growth and stability of the crowdsourcing marketplace, crowdsourcing
W
platforms are a critical linchpin that facilitates the interactions between IT services buyers and
services sellers. Essay three will use a design science perspective to retrospectively evaluate
IE
current crowdsourcing platform designs to assess the viability of these platforms to support the
interests of both services buyers and sellers examined in Essays one and two. Specifically, Essay
EV
three will address the following question: What are the key design principles that are critical to
3
ESSAY ONE
JOINING THE CROWD: FOUR CROW WORKER ARCHETYPES DRIVING
INFORMATION TECHNOLOGY CROWDSOURCING
Abstract
As technology platforms and online communities evolve, the nature of the relationship between
workers and firms is changing. Crowdsourcing is an emerging phenomenon that exemplifies the
changing relationship between workers and firms. Although significant research has been
conducted regarding worker motivation within the traditional firm-worker relationship, relatively
little work has examined motivation in crowdsourcing engagements. This study utilizes revealed
W
causal mapping methodology to identify the motivations and career outcomes of technology
Workforce Survivors, Dream Chasers, Weekend Warriors and Full-time Freelancers. The
PR
findings identify each archetype’s motivations and extend career anchors theory to address new
Introduction
The on-demand-labor economy that uses workers from the general public (crowd) to complete a
task is steadily growing (Bergvall-Kåreborn et al. 2013). This economy rests heavily on the
availability of workers (referred to as crowd workers) who willingly participate in this new
work form of open sourcing mediated by digital marketplaces, which I refer to as crowdsourcing
4
(CS). The popularity of this form of work is evident from the millions of crowd workers who are
registered to work on various CS platforms (Freelancer 2015; Topcoder 2015). The steady rise
of the on-demand or gig economy is drawing attention to the issues surrounding the crowd
workers who participate in this emerging marketplace (Florentine 2015; Kaganer et al. 2013).
Superficially, the lack of job security, the unpredictable work schedules, the lack of a guarantee
of steady income, the absence of insurance and pension benefits through employment, and the
ambiguity regarding safeguards afforded by key labor laws designed to protect the workers
collectively conspire to characterize this kind of work environment as unattractive at best and as
W
a sweat-shop at worst (Kittur et al. 2013). The undesirable attributes of this work environment
prompt a broader question: Why does an individual desire to be a crowd worker? Although this
IE
question has not been examined through the IT crowd workers’ perspective, recently the scholars
have tried to profile this workforce in an effort to understand the individuals who are willing to
EV
participate in this seemingly unattractive work option (Brawley et al. 2016; Rogstadius et al.
2011).
PR
perplexing in the IT labor space, where technology workers often enjoy many employment
privileges due to the talent shortage (Amato 2013). Decreasing numbers of students studying
retirements in the existing technology workforce create a confluence of forces that is expected to
make it difficult for IT organizations to meet staffing requirements over the next ten years
(Bosworth et al. 2013). While companies are increasingly challenged to meet their staffing
needs (Khan et al. 2014; Trost 2014), IT workers are likely to benefit from the low IT
5
meet the demands for IT services. In spite of the availability of good traditional employment 1
and Freelancer continue to attract workers, with over eight million crowd worker accounts
even though crowd workers lack many of the protections available through traditional IT
employment, I posit the following research question: Why do crowd workers participate in IT
crowdsourcing?
W
I address this research question through the perspective of Schein’s career anchors (Schein
IE
1990). Using this theoretical perspective, I categorized why workers participate in
crowdsourcing and characterize differences between the types of individuals who choose to
EV
participate. The career anchors were theorized, in part, to describe worker motivations as
technology diffusion within enterprises began to change work processes. My analysis extends
PR
career anchors theory to the emerging types of worker relationships supported by crowdsourcing.
The remainder of the paper is organized as follows. I next summarize the existing literature on
crowdsourcing and highlight gaps that are addressed in this research. I then present career
anchors theory and use it to examine and describe the motivations of IT crowdworkers. I next
6
Research Background - Types of Crowdsourcing
Crowdsourcing has been defined in a variety of ways. Common themes among crowdsourcing
definitions include utilizing labor from the crowd through an open call to complete a task
mediated or enabled by an online platform (Deng et al. 2016). In some instances the
contributions of the crowd are uncompensated, such as contributors to the website Wikipedia,
and at other times the contributions are compensated, such as through Amazon’s Mechanical
Turk (Doan et al. 2011). While some crowdsourcing research focuses on open innovation
W
et al. 2009), this paper looks specifically at how crowdsourcing may be applied as a means of
support, guidance, and compensation to CS participants (Oshri et al. 2011); I refer to such work
EV
arrangements as IT crowdsourcing (ITCS). Oshri et al. (2011) posited a framework for
and communities. They categorized the types of platforms based on their main focus, structure
of deals, the platform’s role in facilitating buyer-seller interaction, and revenue model (See Table
1).
The “directory” online sourcing environments unite buyers of IT services, or job requestors, and
crowd workers. Directories do not provide transaction processing capabilities, instead serve only
as a virtual job board that facilitates introductions that result in work engagement whose terms
are set outside of the platform. Marketplace environments, which are the focus of this paper,
involve crowd workers’ selecting and finishing posted jobs in exchange for payment.
7
Crowdworkers perform and deliver these jobs according to pre-specified work instructions and a
predetermined compensation rate. The marketplace platform provides tools and support for
framework for engagement and development of crowd workers. Community environments often
structure jobs as contests, where not all the participants are financially compensated. Often only
position themselves as wellsprings of learning where the participants’ visibility and growth are
the salient benefits instead of compensation. By focusing this paper on crowdsourcing for
W
compensation-based marketplaces within the IT sourcing environment, I are able to isolate the
factors that drive IT crowd workers to purse ITCS as a career (i.e., for monetary reasons.) In
IE
these work environments, exemplified by platforms such as e-Lance, Rent-a-coder, Freelancer or
oDesk, job requestors contract with IT workers through the online marketplaces and the
EV
crowdworkers expect to be paid for the delivery of specific services.
While substantial attention has been directed at open innovation contests (e.g., Terwiesch and Xu
PR
2008, Majchrzak and Malhotra 2013; Nevo and Kotlarsky 2014), marketplace crowdsourcing is
limited; for ITCS in particular, it is nonexistent. While some work has been done to examine the
knowledge crowdsourcing as a career pathway specifically within IT has not yet been explored.
from those of less specialized workers. As a larger share of the workforce begins to participate
in ITCS opportunities (Freelancer 2015), it is critical that we develop our understanding of how
to make crowdsourcing arrangements beneficial for both workers and the companies who utilize
8
them. I posit that the implications to workers of crowdsourcing will be particularly acute in
marketplace crowdsourcing work environments, as these workers are participating in jobs for
crowdsourced environments are growing (Kittur et al. 2013). While business leaders may be
intrigued by the possibilities of crowdsourcing and worker advocates may fear the ramifications
of a crowd-based workforce (Schmidt 2013), little work has been done to find out what
W
Table 1. Characteristics of Sourcing Environments IE
Characteristics Directory Marketplace Community
Main Focus Helps buyers discover Connects buyers and Helps members
EV
suppliers suppliers of services develop professional
throughout all stages through interaction
of workflow and paid client
engagements
PR
Structure of Deals are offline, no Deals are done Deals done online
Deals platform involvement online, platform and usually structured
involved in legal and as contests. The
financial aspects of platform is involved
deal in legal and financial
aspects of deal
Platform’s role No buyer-supplier The platform The platform
in facilitating interaction facilitated, facilitates buyer- facilitates buyer-
buyer-seller merely introductions supplier interaction supplier interaction
interaction throughout the with a focus on
project learning and
community building
Revenue Model Advertising, sponsorship Project commission Project commission
paid by suppliers. paid by buyers
Premium
membership fees
9
Platform Infolancer, Chinasourcing Elance, oDesk, Topcoder, Kaggle
Examples Freelancer
Source: Adapted from Oshri et al. 2011
Significant theoretical studies have examined the stages of careers, beginning with the
pioneering work of Super (Super 1957). Super’s work posited four stages of career
corresponding frameworks have been used to define various career stages, including tenure, age,
and organizational level. Another prominent model for describing career stages, presented by
W
Dalton et al., describes four career stages: 1) the apprentice stage, characterized by working
under the supervision of others; 2) the colleague stage, characterized by individual contributions;
IE
3) the mentor stage, characterized by work that can lead to the develop of junior colleagues; and
4) the sponsor stage, characterized by the ability to set direction for the organization. While all
EV
of these models provide a means by which to describe the stages of career progression, they
provide little insight into the work values driving individuals’ transitions through the stages.
PR
Schein's career anchors theory (CAT) (Schein 1990; Schein et al. 1978) provides a theoretical
lens through which to examine motivations that steer individuals’ career pathways. The
overarching thesis underlying the concept of career anchors is that internal values drive an
individual’s career decisions. The types of work values considered career anchors have grown
with additional research regarding worker motivations and interests. In Schein’s early work five
1978). As the relationship between workers and employers has grown more fluid with the
growth of technology and outsourcing, the following four additional anchors have been
10
identified: 6) lifestyle integration, 7) pure challenge, 8) organizational identity and 9)sense of
While the concepts presented by the career anchors framework are applicable to a broad range of
professional endeavors, the framework has been particularly instructive in the examination of IT
careers. IT careers can be marked by rapidly changing skill requirements, and wide ranging
options for career path development. Career Anchors have been used to examine and explain the
career motivations of IS workers and describe the variability of worker expectations that can be
applied to human resource planning (Crepeau et al. 1992). This variability can lead to changes in
W
anchors over time. The age and experience of technology workers has been found to predict a
preference toward security and stability associated with the ability to pursue work opportunities
IE
within a specific geography (Igbaria et al. 1995). The study by Igbaria et al. (1995) also found
that IS experience was positively associated with the career anchor of autonomy. Technology
EV
workers have been found to have dual career options, either pursuing managerial or supervisory
positions, or pursing technical roles (Baroudi 1988). The differing nature of available technical
PR
career paths has led some researchers to propose that workers will align with different anchors at
different stages of their career on different career paths (Agarwal et al. 2000). Chang et al.
(2011) examined how career anchors change over time, finding that technology workers who
positions and in geographic security and value higher levels of autonomy. It is unclear if, and if
so how, these findings apply to ITCS, where the work, career, and organizational boundaries are
more fluid than in traditional employment. By addressing the research question posited in this
11
Career Anchors and the changing nature of Work
While the phenomenon of crowdsourcing continues to grow and a growing number of studies
have examined the behaviors and motivations of crowd workers (Brabham 2010; Deng et al.
2013; Kaufmann et al. 2011), little is understood regarding how IT focused crowdworkers may
Significant theoretical contributions have been made regarding how workers are motivated in
traditional organizations (Amabile 1993; Feldman et al. 1996). Many of these theories address
worker development from the perspective of the employer (Beecham et al. 2008). However, the
W
crowdsourcing work environment puts much greater emphasis on workers managing their own
career development. Career anchors theory provides a lens through which to examine why some
IE
IT professionals participate in ITCS when traditional employment opportunities are plentiful.
While many crowdworkers on platforms like Mechanical Turk may work for far less than
EV
minimum wage (Deng et al. 2013), crowdworkers on compensation-based platforms such as
oDesk may have the potential to earn far more than minimum wage rates. Do crowdworkers on
PR
compensation based platforms see their activities as hobbies or as work? Do they participate in
crowdsourcing to develop skills, for compensation, or both? To the extent that IT crowdworkers
see their involvement in crowdsourcing as “work,” career anchors theory (Schein 1990) can help
uncover the crowd workers’ salient anchors to reveal the work values that motivate their
participation in ITCS in order to provide insight into why they choose to participate in this
Research Methodology
A qualitative approach, revealed causal mapping, is used to effectively evaluate the emerging
crowdsourcing phenomenon and address the underlying research questions, which center on
12
“why” workers participate in ITCS, particularly when traditional employment options are
available. Through the revealed causal mapping approach a robust picture can be developed
of the causes and effects that come through engagement with a phenomena, and the underlying
causes and effects can be analyzed from multiple perspectives. A special issue of
methodologies (Laukkanen 1996; Meindl et al. 1996; Priem 1996), and the methodology was
introduced to the Information Systems discipline through the work of Nelson et al. (Nelson et al.
2000), who expanded causal mapping techniques to a method they called “revealed causal
W
regarding the true motivations and cognitions instigating actions by the linkages between
IE
observed causes and effects among study participants (Fahey et al. 1989).
A revealed causal mapping approach utilizes opened ended questions to elicit feedback from
EV
study participants. While traditional revealed casual mapping is conducted via in-person
interviews, given the distributed nature of technology crowdworkers, I used an online survey that
PR
gives respondents the opportunity to provide text responses. Once data were collected, causal
maps were constructed through analysis of the raw text of survey responses to identify common
themes and commonly related concepts in crowd work motivation and outcomes. The causal
concepts identified in the first wave of analysis were used to create construct-level causal maps
which identified common themes across participants. Raw data were aggregated based on
common themes and concepts, and the relative frequency of each cause and effect relationship
identified in the raw data was calculated. In the study, participants in multiple crowdsourcing
platforms responded to questions regarding motivations and outcomes. The size of the sample
was determined by reaching saturation of cause and effect relationships. Once the point of
13
redundancy in the information was identified, the number of responses was doubled to insure
Revealed causal mapping methodology is well suited to this research question because it allows
for exploratory research into crowdworkers’ motivations while providing a measure of reliability
of the results. The model creates a framework that can be applied in future quantitative work to
Data Collection
I collected data in two steps. First I conducted a pilot study inviting IT crowd workers to
W
complete a survey regarding their work motivations. Invitations to complete the survey were
extended through the online crowdsourcing platforms of oDesk, eLance3 and Rent-a-coder. The
IE
survey was presented as a task and respondents were compensated $20 for their time to complete
EV
the survey. After the pilot data were collected, minor modifications were made to the survey to
better focus on revealing workers’ career anchors and a second wave of data collection was
completed. A total of 25 responses was collected. The survey was available by invitation to
PR
crowd workers, and they were randomly selected from the pool of workers on each platform.
Only workers based in the United States were allowed to take the survey. The online survey
comprised of open ended questions regarding work values and outcomes (such as what types of
jobs they usually take, what they like about doing ITCS, what they dislike, what opportunities
ITCS provides that attracts them to that platform, their career goals, how ITCS helps fulfill their
personal and professional priorities, and whether they would consider doing ITCS as their full
time job) and demographic questions (such as their employment status, age, gender, occupation,
3 The data were collected prior to the announcement that oDesk and eLance would merge.
14