Learned Visual Categorical Perception Effects Depend on Method of Assessment and Stimulus Discriminability

Learned categorical perception (CP) effects were assessed using three different measures and two sets of stimuli differing in discriminability, both of which varied on one category-relevant and one category-irrelevant dimension. Two different kinds of analysis produced patterns of results that depended on both of these variables and show that categorical perception effects are sensitive to variations in assessment task and stimulus discriminability. Only the similarity-rating task produced evidence of between-category expansion effects, suggesting that participants used different strategies for subjective and objective tasks. Generally, there was evidence that category training caused a decrease in the salience of category-irrelevant variation, but when the assessment task cued participants to category-irrelevant differences they were equally apt at identifying categoryirrelevant variation as a control group.