Abstract:
Computational thinking (CT) has been coined a fundamental 21st skill comparable to literacy and numeracy. As a cognitive ability underlying programming and coding skills, CT has been suggested to be fostered early on in education. Accordingly, the last decade significant research effort has been devoted to developing educational activities for CT, integrated into formal and informal educational settings. However, despite the various research on CT, its definition and the respective assessment approaches are still in their infancy. Consequently, the lack of a consensus definition of CT and the limited validated assessment tools for measuring CT restrain the empirical evaluation of the proposed educational materials.
This thesis aims to investigate CT as a cognitive construct and provide an evidence-based definition for it, focusing on the underinvestigated population of elementary school children. To achieve this, first, a review of the literature was conducted to identify concrete CT concepts. Accordingly, a CT curriculum for 3rd and 4th graders was developed, taking into consideration the complexity of the CT concepts and therefore integrating game-based and embodied activities to provide a low-threshold introduction to CT. The novel parts of the curriculum, which are a series of life-size board games, were iteratively evaluated for their usability with adults and children before being integrated into it. After improvements to the games, the curriculum underwent a pilot and an effectiveness evaluation, the latter one designed as a randomized control field trial. In order to measure the effectiveness of the curriculum, a CT assessment tool was developed. Additionally, a correlational analysis was performed in both evaluation phases in order to investigate associations of CT with other cognitive abilities, and therefore complement the nomological network of CT and define the construct.
Results on the effectiveness of the curriculum showed positive effects on students’ CT abilities and therefore appropriateness of the curriculum design, the development and evaluation procedures followed. Moreover, the proposed CT assessment seems reliable for measuring CT at elementary school students and can be used in future studies. Results on the cognitive correlates of CT revealed positive associations of CT with verbal reasoning-, non-verbal visuospatial-, and complex numerical abilities. These results, compared to similar research in other age groups, show similarities but also differences. This implies that CT development is supported by different cognitive abilities across age groups. Additionally, the cognitive abilities investigated in this research could only partially explain CT. This further supports the argument that CT is a specific cognitive ability that builds on and recruits a convolute of several other cognitive abilities, which are not yet extensively investigated in relation to CT.