Uncertainty Monitoring by Young Children in a Computerized Task

Abstract

Adult humans show sophisticated metacognitive abilities, including the ability to monitor uncertainty. Unfortunately, most measures of uncertainty monitoring are limited to use with adults due to their general complexity and dependence on explicit verbalization. However, recent research with nonhuman animals has successfully developed measures of uncertainty monitoring that are simple and do not require explicit verbalization. The purpose of this study was to investigate metacognition in young children using uncertainty monitoring tests developed for nonhumans. Children judged whether stimuli were more pink or blue—stimuli nearest the pink-blue midpoint were the most uncertain and the most difficult to classify. Children also had an option to acknowledge difficulty and gain the necessary information for correct classification. As predicted, children most often asked for help on the most difficult stimuli. This result confirms that some metacognitive abilities appear early in cognitive development. The tasks of animal metacognition research clearly have substantial utility for exploring the early developmental roots of human metacognition.