Perceptual Categorization: Connectionist Modelling and Decision Rules

Although it is currently popular to model human associative learning using connectionist networks, the mechanism by which their output activations are converted to probabilities of response has received relatively little attention. Several possible models of this decision process are considered here, including a simple ratio rule, a simple difference rule, their exponential versions, and a winner-take-all network. Two categorization experiments that attempt to dissociate these models are reported. Analogues of the experiments were presented to a single-layer, feed-forward, delta-rule network. Only the exponential ratio rule and the winner-take-all architecture, acting on the networks' output activations that corresponded to responses available on test, were capable of fully predicting the mean response results. In addition, unlike the exponential ratio rule, the winner-take-all model has the potential to predict latencies. Further studies will be required to determine whether latencies produced under more stringent conditions conform to the model's predictions.     

Mobilization of Cognitive Resources and the Generation Effect

The generation effect refers to the memory advantage of words that have been generated rather than read. Such a read-generate comparison confounds qualitative task differences and raises methodological problems. A revised methodology is proposed circumventing these problems in that the encoding task is held constant and all stimuli have to be generated, but the degree of generativeness (i.e. the amount of cueing) is varied. In Experiment 1, 1, the (refined version of the) generation effect is demonstrated in a within-subjects design; with increasing generation activity left to the subject, free recall performance increases. No effect is obtained for degree of target masking. The same finding is replicated and shown to be independent of self-paced study time when generative activity is manipulated between subjects (Experiment 2) or within subjects (Experiment 3). As all learning trials involve generation, encoding time is controlled statistically, and free recall is used as a measure of memory, this refined generation effect cannot be explained as an artifact of selective attention or elaboration. Rather, generative activity seems to increase the mobilization of cognitive resources. This motivational account is supported by Experiment 4 showing an enhanced generation effect for positive mood.