Memory and representativeness.

We explore the idea that judgment by representativeness reflects workings of memory. In our model, probability a hypothesis conditional on data increases in ease with which instances are retrieved when cued data. Retrieval is driven measure similarity exhibits contextual interference: data/cue less likely to retrieve occurs frequently other As result, assessments context dependent. new laboratory experiment, participants shown two groups images different distributions colors and features. line model's predictions, we find (a) decreasing frequency given color one group significantly recalled group; (b) cueing features for same set entails probabilistic assessments, even if normatively irrelevant. A calibration model yields good quantitative fit data, highlighting central role interference. (PsycInfo Database Record (c) 2020 APA, all rights reserved).