Exploring the influencing factors of computational thinking of primary school students

Abstract

Along with the advancement of information technology, computing skills are increasingly required in many aspects of society, and computational thinking (CT) has been considered an essential component of 21st-century skills. The importance of CT has been highlighted through the growing development of CT education. In recent years, cumulative efforts were seen in teaching CT at gradually earlier educational stages. However, limited was known regarding the factors underlying children’s CT acquisition. Therefore, this thesis aimed to explore the influencing factors of CT of primary school students. To achieve this goal, three overarching research objectives were proposed, involving developing a robust CT instrument, exploring cognitive influencing factors, and exploring noncognitive influencing factors. Three studies were administered to address these objectives respectively. Study 1 developed a CT instrument for primary school students (aged 6-11). A principled approach was leveraged to support the design process, and systematic validation procedures were performed to verify the plausibility of the measurement. The results reported adequate validity and reliability properties, indicating the appropriateness of the instrument for the target group, which supports the measurement of CT in the next two studies. Study 2 investigated the cognitive correlates of CT. The associations between CT and specified cognitive domains were examined, namely, mathematical, spatial, and reasoning skills. To improve the robustness of the results, fine-grained skills of each domain were focused on, and the effect of potential confounding variables was considered. The results indicated that students’ CT acquisition can be positively predicted by their performance in numeracy, spatial perception, spatial visualization, and nonverbal reasoning. Study 3 investigated the noncognitive correlates of CT. The associations between CT and specified noncognitive factors were examined, involving demographics, prior experience, attitudinal beliefs, and personality. To reveal the underlying mechanism of the CT-noncognitive relationship, the study explored the intermediary process that bridged personal features with CT acquisition. Results revealed that better coding attitudes and greater conscientiousness contribute to higher CT performance. Also, the mediating role of coding attitudes between personal features (i.e., gender, coding experience, and personality traits) and CT was detected, where students who were male, more experienced in coding, more open-minded, and less negatively emotional tended to perform better in CT via the effect of better coding attitudes. The study contributes to the development of CT primary education by (1) providing a robust CT instrument with replicable development and validation procedures, (2) enriching the conceptual understanding of CT from both cognitive and noncognitive lenses, and (3) proposing a nomological network of CT to implicate the integration of these factors into teaching practices. Methodological, theoretical, and practical implications to the field are presented, and limitations and future directions are discussed.