OVERESTIMATION OF BASE-RATE DIFFERENCES IN COMPLEX PERCEPTUAL CATEGORIES

The optimality of multidimensional perceptual categorization performan ce was examined for several base-rate ratios, for both integral and se parable dimension stimuli, and for complex category structures. In all cases, the optimal decision bound was highly nonlinear. Observers com pleted several experimental sessions, and all analyses were performed at the single-observer level using a series of nested models derived f rom decision-bound theory (Maddox, 1995; Maddox & Ashby, 1993). In eve ry condition, all observers were found to be sensitive to the base-rat e manipulations, but the majority of observers appeared to overestimat e the base-rate difference. These findings converge with those for cas es in which the optimal decision bound was linear (Maddox, 1995) and s uggest that base-rates are learned in a similar fashion regardless of the complexity of the optimal decision bound. Possible explanations fo r the consistent overestimate of the base-rate difference are discusse d. Several continuous-valued analogues of Kruschke's (1996) theory of base-rate learning with discrete-valued stimuli were tested. These mod els found some support, but in all cases were outperformed by a versio n of decision-bound theory that assumed accurate knowledge of the cate gory structure and an overestimate of the base-rate difference.