关键词: 基于类别的特征推理, 特征联结, 预测特征, 竞争范式

Abstract: One of the most important functions of category was to predict unknown features. But when categorization was uncertain, how did people predict? Based on categories or based on feature associations? A new paradigm in which there were two competitive predicted features was used to investigate it. In the new paradigm, there were two to-be-predicted features; the feature probability of Feature 1 was higher than the Feature 2. If people made predictions based on categories, they would make a higher probability judgment for Feature 1. For feature associations with known features, Feature 2 was higher than Feature 1. If people predicted based on feature associations, they would make a higher probability judgment for Feature 2. There were two experiments; Experiment 1 was interested in how people predicted when categorizations were ordered before prediction. Experiment 2 explored people made predictions based on categories or feature associations when categorizations weren’t ordered before prediction. The results found that when categorizations were ordered participants predicted based on categories, while categorizations were not ordered, people inclined to predict based on feature associations. The results indicated that people make predictions in an economical cognition processing approach.

Key words: category-based feature prediction, feature association, predicted feature, competitive paradigm