Influence of altered goal-box cues on partial-reinforcement effect.

Abstract

The purpose of the experiment was to study theoretical interpretation of the partial-reinforcement effect (PRE). Amsel proposed that nonreinforcement during training generates frustration. Weinstock and Hulse stressed the importance of habituated competing responses and strengthened competing responses learned to the cues of the empty goal box. Assessment of the positions, however, is difficult, since nonreward during training is adequate for the development of both the habituation and the strengthening of competing responses as well as the development of frustration reactions. An attempt was made to separate the effects of frustration from those of competing response positions by reducing during extinction the emission of learned responses in the goal box but not in the start box and runway. Since the assumption is made that the PRE is dependent on learning that occurs in the goal box, the competing-response positions would predict a radical change in the goal box during extinction to inhibit the PRE. Since the frustration position asserts that the PRE is a function of learning in the goal box that generalizes to all sections of the apparatus, the PRE should occur to some degree despite altered goal-box conditions. Using a straight runway apparatus, subgroups of rats trained on 100% and 50% schedules of reinforcement were extinguished to either familiar or novel goal-box conditions. Subgroups trained on 100% and 50% schedules of reinforcement exhibited a significant PRE under both the familiar and altered goal-box conditions. Findings were discussed as supportive of the frustration hypothesis.