Exploring Control in Early Computing Education

In the paper we reflect on how our design research approach in the current development allows us to study the increasing cognitive complexity of different levels of control which pupils conduct when they program Emil, a virtual character on the screen. In our earlier work we outlined conceptual framework for primary programming, which recognised three different levels of control: (a) direct manipulation, (b) direct control and (c) computational control (i.e. programming) an actor. In the present research we managed to get deeper into the complexity of control by identifying four instead of three of its levels. Based on our close collaboration with three design schools we have also found that it is more productive to project and analyse learning progression of pupils connected with control within two-dimensional grid, where the first dimension is control itself and the second explores the way how the control is represented. Along this dimension we have identified five distinct levels of representation: (a) none, (b) as internal record, (c) as external record, (d) as internal plan for future behaviour, and finally (e) as external plan for future behaviour. In our paper we explain the grid of control by presenting selected tasks from different environments of Emil, our new approach to educational programming for Year 3 pupils.

1. 1.

   Here we borrow from [2] the dimensions of their computational thinking framework, however as we explained in [3] we prefer to broaden the dimension of computational concepts into computational constructs, i.e. concepts plus associated computational procedures (e.g. a sequence of steps as a concept and acting it, interpreting, filling in a missing step, comparing two sequences, modifying a sequence etc. as some of related computational procedures); and also, we consider control – the way how pupils give orders to a sprite or a programmable toy – to be one of the key Brennan’s and Resnick’s computational practices.

2. 2.

   Aged 5 to 10.

3. 3.

   By suitable steps we mean gradations of tasks which support all pupils in exploring these constructs and practices and constructing their true and sustainable understanding. We strive to do so despite the fact these concepts and practices are often wrongly considered trivial.

4. 4.

   We deliberately narrowed Blackwell’s view from all non-professional programmers to pupils.

5. 5.

   With occasional deflections, see our comment on controlling Bee-Bots later in the paper.

6. 6.

   There is a deflection though from basic direct drive strategy in Bee-Bots. If we want to give it a single command then run it, we have to press an arrow key, then press GO, then before the following command is pressed, Clear the memory. Otherwise the next command would be added at the end of the previously recorded steps. This makes direct drive with Bee-Bots less straightforward and we in our Bee-Bot pedagogy recommend advancing from direct manipulation to incremental recording of the program, as described above.

7. 7.

   Although we consider it developmentally appropriate affordance of Bee-Bots.

8. 8.

   Note that such dragging (we call it ‘meta dragging’) is in Scratch indicated by a shadow rim around a sprite.

9. 9.

   Having analysed and explored many alternatives.

10. 10.

    In its current, not final state as this is on-going process. We comment on this issue in the closing remarks.

11. 11.

    We refer to them as the design schools, see more details in [11].

12. 12.

    Currently the lessons are already run by the class teachers themselves, with our continuous support.

13. 13.

    i.e. planning its future behaviour.

14. 14.

    As a preparation for later perciving programs to be objects to think about and think with.

15. 15.

    Including Scratch, as explained in 2.2 How we control in Scratch.

16. 16.

    Blackwell does not distinguish between direct manipulation and direct control.

17. 17.

    Which in the case of virtual actor would correspond to what we call ‘meta dragging’ by mouse.

18. 18.

    Which regularly led to modifying or reorganizing the tasks, adding new ones and removing others, transforming a task into a whole new unit of tasks.

19. 19.

    We sometimes refer to this as ‘meta dragging’.

20. 20.

    with several constraints, which pupils will eventually discover by exploring – no diagonal clicks are allowed, no clicks behind a missing possition and several other constraints.

21. 21.

    As we discuss later, the second dimension gives different categorisation of the tasks then the control dimension by itself. When we use both dimensions in one grid, see Fig. 6, each of the resulting 20 combinations (positions) has a meaningful interpretation in educational programming.

22. 22.

    In [11] we refer to it more precisely as a pre-construct.

23. 23.

    To start in October 2018.

The authors would like to thank Indicia, non-for-profit organisation funding our project, all the teachers and pupils from our design schools for their invaluable contributions to the design and development of Emil intervention, and Celia Hoyles, Richard Noss and James Clayson for exciting discussions about the issue of control in educational programming.

Authors and Affiliations

1. Comenius University, 842 48, Bratislava, Slovakia

   Ivan Kalas & Andrej Blaho

2. Edix, Bratislava, Slovakia

   Milan Moravcik

Corresponding author

Correspondence to Ivan Kalas .

Editors and Affiliations

1. Saint Petersburg Electrotechnical University, St. Petersburg, Russia

   Sergei N. Pozdniakov

2. Vilnius University, Vilnius, Lithuania

   Valentina Dagienė

Reprints and permissions

© 2018 Springer Nature Switzerland AG

Policies and ethics