Memory-resource-allocation experiment in rats: Failure to find directed forgetting effect

Abstract

Three rats were trained in a directed forgetting task using an eight-arm radial maze. A trial consisted of learning phase, retention interval, and test phase. In a learning phase, rats were allowed to get food A at two randomly-decided arms and food B, also at two randomly-decided arms. No food was set at the remaining four arms. During the 3 min retention interval, rats were kept in either a large transparent cage or a small metallic cage. In a test phase, food rewards were set at arms that had been rewarded in the learning phase (win-stay task). Only one of the food rewards, either A or B, was tested. Doors of the arms where the other rewards were set in the learning phase were closed and excluded from the test. Which arms would be tested was signaled by the type of holding cage in which the rats were kept during the retention interval. After rats’ performance was reached around 80% correct, probe test trials were inserted in the training trials. In a probe test trial, those arms that were inconsistently signaled by the holding cage were then presented in the test phase. Superior performance was expected in training trials compared to the probe trials, if rats could reallocate memory resource for arms not tested to ones to be tested. As a result, performance of training trials did not exceed that of probe trials, suggesting no directed forgetting effect. Rats’ learning strategy and methodological problems in the present experiment are discussed.