Decision-boundary theories of categorization are often supported by results from experiments that use category structures which make boundary models difficult to distinguish from supported by results from experiments that use category structures which make boundary models difficult to distinguish from exemplar-based models. We present two experiments that potentially differentiate the two model types. The results of both experiments point strongly to the absence of decision boundaries in most participants. No participant displayed use of the optimal boundary. The range of non-optimal solutions shown by individual participants is accounted for by an exemplar-based adaptive learning model.
Available electronically via the Web at
http://www.indiana.edu/~kruschke/kalishkruschke_abstract.html