Proceedings of the Annual Meeting of the Cognitive Science Society

Efficient algorithms can enhance problem-solving in many cognitive domains but can be difficult to discover and use. For example, classical studies of seriation suggest that children struggle to apply algorithmic strategies in a simple sorting problem. We investigate the spontaneous discovery of algorithmic solutions across development. We gave children a variant of the sorting problem with hidden object ranks. Children sort animated bunnies into the right order, from the shortest to the tallest, when the bunnies are standing behind a wall so their heights are not visible. Children performed far above chance on this difficult sorting task, potentially because higher demands in memory and reasoning incentivized strategic behaviors. Children independently discovered at least two efficient algorithmic solutions to the sorting problem, Selection sort and Shaker sort. This result suggests that children are far more competent at sorting tasks than previous research would suggest. Additionally, older children were more efficient sorters than younger children. This suggests that competent performance on sorting tasks improves throughout development.