Assessing Cognitive Components of Computational Thinking

Andrew Joseph Mertens ^* Eliana Colunga

• University of Colorado Boulder, Boulder, United States

Recent years have seen a dramatic increase in Computer Science (CS) education programs implemented at the K-12 level. This emphasis on CS education comes not only from the fact that computer skills are becoming an ever-more integral part of modern life, but also from a notion that learning how to program facilitates the development of a set of more general skills and strategies collectively known as Computational Thinking (CT). What makes CT special is the fact that it can be applied in an array of different contexts that are not limited to the CS domain.The present work adapts tasks from established cognitive tests in an attempt to capture some of the components specifically theorized to comprise CT, namely decomposition, sequencing, and abstraction. To test this, we conducted three studies to test the degree to which these measures relate to proficiency and experience with computer programming. Study 1 examines this relationship in 8-12 year-old children enrolled in STEM summer camps. Study 2 examines the programming proficiency-CT relationship in a different population and setting: fourth graders in a public elementary school. Study 3 aims to contribute converging evidence for the relationship by measuring CT and programming experience in an online study in the 8-12 year-old STEM summer camp population. The results reveal that performance on the decomposition measure consistently correlates with both proficiency and experience measures of programming in young children. We discuss these findings in the context of a potential progression for the emergence of CT-related skills throughout development.