The effects of feature analysis, transparency in probabilistic category learning in adults and children

Abstract

Abstract The Weather Prediction (WP) Task is one of the most widely used tasks in probabilistic category learning research. Earlier studies mainly tested adults on similarly structured but different versions of the WP task, mostly without specific focus given to the differences between these tasks. The current paper focuses on the effects of stimulus organization on learning, manipulating two variables: 1) transparency of cues and outcomes and 2) combination of cues into features of a single image vs presenting them as distinct cues. Results show that different variables affect probabilistic categorization differently; cue-based (as opposed to holistic) presentation leads to better performance, and transparency also helps categorization in the early phases of the task, but this advantage turns into a disadvantage later. In general, adults’ probabilistic categorization abilities surpass children's categorization, but the pattern of performance was similar regardless of age.