Frustration Tolerance and Incidental Learning as Determinants of Extinction

Abstract

Two theories currently exist to explain the retarding effect of partial reward on the extinction of a learned behaviour. Amsel^1,2 has suggested a theory based on the thesis that the shortcomings (non-reward) of a partial reward schedule generate frustration, experience of which reduces the reaction to the frustration induced by extinction. We shall here refer to this as the “frustration tolerance” theory. Sutherland^3,4, on the other hand, has proposed a theory based on the assumption that partially rewarded animals switch attention more often than consistently rewarded animals and therefore learn about more cues. During extinction the responses to all the cues which the animals have learned about during training have to become extinguished. This suggests that the more cues learned about the longer it takes to extinguish the responses to them. We shall refer to this as the “incidental-learning” theory.