类别学习中分类学习与推理学习的对比研究

A Comparalive Study of Inference Learning and Classification Learning in Category Learning

本研究对类别学习中分类学习与推理学习进行了对比。实验中要求被试从成对呈现的项目中选择一个更典型的A,并且要求被试将典型的A和典型的B画出来。测试中所呈现项目诊断性程度或典型性程度不同。研究结果表明,分类学习者主要受诊断性信息的影响,而推理学习者主要受典型性信息的影响。即,分类学习者关注类别间信息,而推理学习者关注类别内信息。

In the contrast of classification learning and inference learning, we presented pairs of testing stimuli and asked the participants to select a more typical A from them, and then asked the participants to draw pictures of a typical A and a typical B. These testing stimuli were different in diagnosticity or prototipicality. According to the results, we conclude that classification learners are influenced primarily by diagnotic information, while inference learners are influenced much more by the prototypical information. In other words, classification learners focus on between-category information, and inference learners focus on within-category information.