Discrimination reversal training with single and multiple stimulus pairs in the squirrel monkey.

Abstract

2 groups of squirrel monkeys were used to study discrimination reversal with a single pair vs. multiple pairs of problem objects. The Single Pair group required twice the number of trials that the Multiple Pair group required to achieve successively the criterional performance of 6 consecutive correct responses. Subsequently, both groups were given criterional discrimination reversal problems with multiple problem pairs. As both groups performed equally throughout this phase, earlier differences were interpreted as reflecting a performance variable rather than increased error factors in the Single Pair condition. Terminal-test performance for combined groups was 88% correct on prereversal and 82% correct on reversal trials.