Social belief revision in the form of advice-taking is well studied in adult social psychology. Adults do take others' advice systematically and selectively, but are far from behaving optimally, even if that means a loss of decision accuracy. The development of selective belief formation (selective trust and learning) has been intensively investigated in recent years, yet its cognitive basis remains unclear. In addition, basically nothing is known to date about the development of selective belief revision in children. Our project contributes to closing this gap by developing a standard judge-advisor-paradigm that is applicable both to children and to adults.

Children played a computerized game in which they were to assign the right amount of food to fishes with different features (stripes, number of fins, size). It was made clear to children that both feeding too little (‘‘then the fish remains hungry’’) and feeding too much (‘‘then the fish gets a stomach ache’’) would have negative consequences for the animal. After they made their initial judgement by touching the “feeding machine” children received advice from another child who played the same task in a prior study. In the end, they were invited to make their final decision.

The results so far indicate a quantitative difference in advice-taking behavior between children and adults (i.e., higher mean AT scores of children compared with the usual level of advice taking observed in adults).