Computational thinking through programming: a conceptual model for teaching and learning computational thinking

Abstract

Computational Thinking (CT) through programming attracts increased attention as it is considered an ideal medium for the development of 21st century skills. CT initiatives have emerged around the world and there is a rapid increase in relevant research studies. The accumulation of research plethora leads to the need for a conceptual model of CT that could map the domain, facilitating comprehensive understanding of the domain’s challenges. The aim of this thesis is a) to develop a conceptual model based on a systematic literature review that maps CT through programming in K-12 and higher education and b) to investigate the relationships between certain instances of the model, namely of the effects of scaffolding programming games and attitudes towards programming, on the development of students’ Computational Thinking. Regarding the first aim of this thesis, the proposed Computational Thinking through Programming in K-12 education (CTPK-12) conceptual model emerges from the synthesis of 101 studies and the identification of CT Areas. The proposed model consists of six CT Areas (namely Knowledge Base, Learning Strategies, Assessment, Tools, Factors and Capacity Building) and their relationships. Some of the relationships between areas have not yet been sufficiently explored in the literature such as which learning strategies enable the development of CT. The revised model for higher education is derived from a systematic mapping of 41 studies. This model includes the same CT Areas and relationships as CTPK-12, however it differs in sub-areas and instances. Knowledge Base Area, Assessment Area and Tools Area have significantly evolved throughout the years, while Capacity Building Area has only recently emerged. In addition, the introduction of CT to undergraduate students and preservice teachers differs mainly in the tools used and the CT elements that are assessed. Regarding the second aim, students were introduced to CT under two distinct experimental conditions: a scaffolding version of a programming game and a non-scaffolding version of the same game. Results report statistically significant differences between the pre-intervention and post-intervention CT scores for all students and statistically significant improvement in learning outcomes in favor of the scaffolding group. In addition, the study hypothesized that attitudes towards programming would have an impact on students’ CT. Although this hypothesis has not been confirmed, the results suggest that students who have a less positive attitude towards programming could particularly benefit from scaffolding aspects in programming games.