Abstract
Predetermined sequences of stimuli and a direct measuring technique were used to estimate the specific hypotheses Ss used to solve concept-identification problems. Results indicate that Ss did not choose hypotheses at random and that the focus sample of most Ss contained those hypotheses which were consistent with reinforcement on previous trials. A comparison was also made between predictions of the Trabasso and Bower (1968) and Chumbley (1969) models concerning Ss' sampling behavior after error trials. Ss eliminated irrelevant hypotheses on both error and correct choice trials at about the same rate, supporting the predictions of the Chumbley (1969) model.
Get full access to this article
View all access options for this article.
