Previous studies indicate that patients with damage to the striatum, such as patients with Parkinson's disease (PD) or Huntington's disease (HD), are impaired in certain categorization tasks, but show no impairment in other categorization tasks. These studies suggest that the striatum may be involved in category learning under some circumstances but not others. One possible role of the striatum in category learning is that these structures are involved in learning nonverbal rules, but only when learning is based on corrective feedback under supervised learning conditions. Such a hypothesis is consistent with current models of striatal functioning. However, it is difficult to draw strong conclusions regarding the proposed role of the striatum based on past work because most of these studies used very different categorization tasks that vary along a number of important dimensions. The proposed research remedies these problems by conducting highly systematic studies of category learning in patients with PD. Factors known to impact the verbalizability of categorization rules will be explored, including (1) whether the rule is linear or nonlinear, (2) whether the rule requires information integration across dimensions or selective attention, and (3) whether the stimulus dimensions are separable or integral. In addition, the nature of training (corrective feedback or observation) will also be examined. These factors likely determine the extent to which the striatum is involved in category learning. Each of these factors will be explored within the framework of a highly successful categorization paradigm that has been used extensively in studies of normal cognition, and recently has been extended to some patient populations and normal aging. The paradigm, called the perceptual categorization task, is rigid enough that strong controls can be placed on factors that vary widely across other tasks, but is flexible enough that each of the factors outlined above can be studied in isolation. Further, quantitative models will be applied to the data of PD patients and controls in order to determine more precisely the nature of any observed category learning deficits in the PD patients.
Showing the most recent 10 out of 16 publications
