Verbal mediating responses: effects on generalization of say-do correspondence and noncorrespondence

We analyzed the effects of verbal mediating responses on the acquisition and generalization of say-do correspondence and noncorrespondence. Participants were assigned to groups in which either reinforcers (feedback and tokens) were arranged for say-do correspondence and noncorrespondence, or no reinforcers were programmed. Participants in these groups were further subdivided into groups in which they were required to repeat what was said previously, were required to repeat random numbers, or no verbalizations were required. When correspondence was reinforced, repetition of what was said produced greater acquisition and generalization of correspondence. When noncorrespondence was reinforced, repetition of numbers facilitated acquisition and generalization of noncorrespondence. Verbal mediating responses interacted with contingencies of reinforcement in determining acquisition and generalization of correspondence and noncorrespondence.