Goal-directed or open-ended? Investigating children's tool innovation across contexts

Although humans are uniquely prolific tool users, children consistently struggle to innovate tools in experimental paradigms, a finding that challenges perceptions of children as natural creative beings. This paradox may reflect limitations in experimental paradigms that impose rigid structures misaligned with children's natural, exploratory problem-solving tendencies. The present study examined whether more open-ended contexts better reveal children's capacity for tool innovation and explored the processes underlying creative problem solving. A total of 126 British children (66 girls, 60 boys, 0 nonbinary), aged 4-7 years (M = 6.04, SD = 0.94), were tested in dyads on two tool innovation tasks, each presented in either a goal-directed or open-ended condition using a counterbalanced design. The tasks were themed to encourage play. In the pirate-themed hook task, treasure was available to retrieve from a hole on an island by creating a pipe cleaner hook, and in the alien-themed floating-object task, a mysterious object was available to retrieve by filling a tube with water to make the object float. Innovation rates did not differ across the goal-directed and open-ended conditions for either task. However, further analysis revealed that, regardless of condition and age, children who engaged in more exploration were more likely to innovate successful tool solutions. These findings underscore the central role of exploration in children's tool innovation and highlight the importance of examining problem-solving processes beyond outcome-based measures. (PsycInfo Database Record (c) 2026 APA, all rights reserved).