Children’s Behavior to Response-Independent Stimuli: A Replica of Wagner and Morris’ (1987) Superstition Experiment

Skinner (1948) dubbed the idiosyncratic responses of pigeons exposed to noncontingent accidental or adventitious reinforcement “superstitious”. According to Skinner’s interpretation, the rate of these responses is expected to be higher under richer, non-contingent schedules (e.g., FT15s) than under leaner schedules (e.g., FT30s). However, children’s response rate in replications of Skinner’s superstition experiment did not show a clear difference between FT15s and FT30s schedules. This study evaluated whether the response rate was higher during an FT15s or an FT45s schedule by replicating Wagner and Morris’ (1987) methodology. Five-year-old children played with a tablet application in which they received virtual balls independently of their behavior. We expected that the response rate of idiosyncratic responses displayed during the game should have been higher under the FT15s schedule than under the FT45s, as accidental reinforcement frequency is higher. Our findings did not conform to Skinner’s interpretation. An approach based on stimulus control and behavioral systems better accounts for our results.