ABSTRACT
This conclusion presents some closing thoughts on the concepts covered in the preceding chapters of this book. The book argues that, according to the independent cue models, categorization learning should be easier for the separable than for the nonseparable condition in this experiment. It provides a thorough review of the relevant evidence. A major problem running through all the research is the difficulty of ascertaining that techniques designed to detect a learner’s mode of processing can be used without, in the process, disturbing the course of learning. Several modifications of the original model were proposed to enable it to accommodate findings of perfect discrimination of similar stimuli. The family of hypothesis-selection models has provided instructive interpretations of aspects of conceptual learning by human beings and other higher primates that seem, for the present at least, to be beyond the reach of the learning models.
