by Luzia Leifheit, Julian Jabs, Manuel Ninaus, Korbinian Moeller, and Klaus Ostermann

In European Conference on Games Based Learning (ECGBL). "Academic Conferences and publishing limited", 2018.

This publication is related to the Computational Thinking research project.

Abstract

Game-based approaches can be a motivating and engaging way of teaching and learning, in particular for younger students. To evaluate the suitability of game-based approaches for teaching programming in primary school, we conducted a field study on Computational Thinking (CT). CT can be characterized as the ability to understand, formulate, and systematically solve complex problems, which typically requires abstraction, generalization, parameterization, algorithmization, and partitioning as processes of CT, which are also vital to programming. The employed CT course focused on fostering students’ conceptual understanding of computational thinking independent of specific technological applications and was based on course material from code.org. Lessons primarily addressed algorithms as a core CT concept and used game-based learning material to increase students’ understanding of algorithmic CT concepts such as sequences, loops, branches, and events. These concepts were first introduced through unplugged game-based activities using tangible everyday objects (e.g. pencils, playing cards, etc.) instead of abstract code. In more advanced lessons, students’ conceptual understanding was applied and deepened through plugged-in programming exercises. The 18 sessions (45 minutes each) of the course were taught to 33 3rd and 4th grade primary school students. At the end of unplugged lessons, students’ understanding of newly introduced CT concepts was assessed by short tests. Students’ interest in and motivation for programming education was measured with pre- and post-course self-assessment questionnaires. Results indicated benefits of the unplugged game-based approach for teaching CT concepts. In particular, we observed (a) for all but one of the CT concepts, students reached on average 82% of the learning objectives or more, and (b) students rated their learning experience in the course positively and reported high levels of interest in learning more about computation-related topics. In addition, qualitative analyses indicated part of the curriculum was very complex for the target group (e.g. nested conditionals). This finding is of particular interest for the development and evaluation of future programming courses for primary school students.