Christopher G. Prince,

Center for Advanced Computer Studies, USL

&

Kelly A. Kimball

Departments of Psychology & Anthropology, USL

(Presented at the Sigma Xi First Annual USL Student Research Symposium, April 28th, 1997)

Experiments are being conducted at the New Iberia Research Center to determine if rats' cognition includes learning relations between locations. Since previous research (e.g., Macuda & Roberts, 1995) has provocatively suggested that rats are able to learn hierarchically structured groupings of locations, we are employing different but related methods to provide further evaluation of this hierarchical structure (HS) hypothesis. Using plus maze configurations (e.g., NORTH, SOUTH, EAST, WEST) where maze arms are respectively designated as X1, X2, Y1, and Y2 rats are being given trials in which the starting arm on a trial is "X" and the single baited arm is "Y." For example, one rat receives trials in which WEST is the start arm, with the NORTH arm baited, and trials in which SOUTH is the start arm, with the EAST arm baited. Each of the 12 rats in this study receives 8 trials/day where 4 trials are "if X1 then Y1" (constituting a left turn) and 4 trials are "if X2 then Y2" (constituting a right turn). One rat to date has met the learning criterion (28/32 on both trial types across 8 sessions), achieving correctly 30/32 and 29/32 on the two trial types; two other subjects have statistically significant correct performance (p < .0001) and appear close to meeting this criterion. A "symmetry" test, in which the contingencies are reversed, is being conducted with rats that meet the learning criterion. In this test the "Y" arm is now the start arm and the "X" arm is now baited. While there are a number of methodological issues yet to be addressed, the degree to which the HS hypothesis is valid suggests the degree to which, after learning the "if X then Y" relations, the animals should show immediate performance on the "if Y then X" relations.