Learning Set with Intra-Problem Stimulus Variation

Abstract

Thirty-six discrimination problems of a new type were presented to teen-age human Ss. Each problem contained four trials, and each trial involved a choice between a constant and varied stimulus. The constant stimulus appeared on all trials within a problem The varied stimulus was different on every trial. There were six experimental groups, representing orthogonal combination of a three-treatment reinforcement variable and a two-treatment response vs nonresponse variable. The reinforcement variable involved the following treatments, listed in ascending order of difficulty: constant stimulus correct on all problems, varied stimulus correct on all problems, constant and varied stimuli correct alternately over problems. The response vs nonresponse variable was applied only to Trial 1, and its effects were evaluated on later trials, which showed that the nonresponse treatment was more difficult.