抄録
Two experiments were performed with 96 undergraduates to examine recognition memory of average prototypical exemplars, modal prototypical exemplars and original exemplars in a repeated study-recognition paradigm. In Experiment 1, the amount of within-category variability and the instruction for memorizing were varied. In the narrow category-structure which had less-variable exemplars, the confidence scores were higher for the average prototypical exemplars than for the modal ones, whereas the reverse was true in the wide category-structure which had morevariable exemplars. The confidence scores of modal prototypical exemplars were higher under the abstraction instruction than under the exemplar one. With repeated study-recognition trials, the confidence scores increased for the original exemplars but decreased for the average and modal prototypical exemplars. In Experiment 2, the amount of within-category variability and the presentation procedure of original exemplars were varied. Although the results were about the same as in Exp. 1, the confidence scores of modal prototypical exemplars were higher when the original exemplars were presented successively than when they were simultaneously. The findings were interpreted with reference to distinctiveness of dimensional values of the original exemplars and discriminability between the original exemplars and the prototypical exemplars.
