No abstract

No abstract

The abstraction of prototypical information and the classification of new exemplars was investigated as a function of category size and the number of categories that had to be distinguished during learning. It was argued that category size should influence primarily the abstraction of information common to that category, whereas the number of categories that had to be distinguished during learning should provide for the isolation of distinctive features among the categories. The results indicated that when four or six categories were distinguished, category size (three versus six exemplars per category) had a facilitative effect on the classification of the prototype and new exemplars; the effect of category size was minimal when there were two categories. In addition, conversion to d' values revealed a slight main effect of the number of categories on prototype abstraction. It was concluded that both variables are operative in the abstraction from ill-defined categories, but that the influence of common and distinctive features may occur at different stages of learning. A secondary result was that rapid learners were also superior abstractors, although even the slowest learners benefited by category size.

No abstract

No abstract