Proceedings of the Annual Meeting of the Cognitive Science Society

Computational thinking (CT) refers to a range of problem-solving skills applicable to computer science and everyday life.Although recent research in developmental cognitive sciencesuggests mental capacities relevant to CT may emerge quiteearly in life, research on CT, and computer science educa-tion more generally, has made little contact with this litera-ture. As a way to better bridge these fields, we explore thedevelopment of problem decomposition, a critical feature ofCT, in the spatial domain. We ask whether young childrencan break a complex spatial problem down into subcompo-nents that can be reassembled to solve the overarching prob-lem. Across two experiments (Exp.1: 4- to 7-year-olds; Exp.2:3- to 5-year-olds) that involve constructing block structures,we demonstrate that some of the key capacities underlyingproblem decomposition begin to emerge in preschool years anddevelop throughout early childhood. Although preschool-agedchildren struggle to solve an open-ended decomposition prob-lem that requires generation and execution of decompositionplans, even 4-year-olds can successfully evaluate the viabilityof these plans. These results suggest that experimental meth-ods in developmental cognitive science can inform CS edu-cation research that focuses on promoting CT; by identifyingwhen and how CT concepts emerge in early childhood, we canbetter create age-appropriate educational tools.