The Differential Effects of Stimulus Presentation during Error- and Success-Feedback Intervals in Concept Identification
MICHIGAN UNIV ANN ARBOR HUMAN PERFORMANCE CENTER
Pagination or Media Count:
To test an hypothesis that information is processed better after successes in concept identification than after errors because of failure to remember the stimulus, 40 Ss were run in 16 simultaneous-discrimination problems. Blank trials were included to show Ss hypothesis after each error. In group ES the positive member of the stimulus pair was presented during the feedback interval on all feedback trials. This was done only after errors in group E, only after successes in group S, and on no trials in group C. Ss responses between errors were almost always consistent with exactly one hypothesis. Ss in groups E and ES were more locally consistent, more globally consistent, and more likely to choose the correct hypothesis after errors than Ss in groups S and C, consistent with the idea that failures to process efficiently after errors are due to memory loss. Ss in all groups showed more global consistency than expected on the basis of random or just locally consistent sampling by Ss. Modifications of several models of concept identification were developed in the light of these results.