Determinants of spaced-trial paradoxical learning effects in pigeons

Abstract

Three experiments explored spaced-trials paradoxical learning in pigeons. Experiment 1 compared the effects on persistence of partial reinforcement to continuous reinforcement in a runway task at one trial per day. Pigeons on partial reinforcement demonstrated greater persistence than subjects on continuous reinforcement when reinforcement or trials were equated. Experiment 2 compared persistence on a variable magnitude of reward schedule to those of a continuous reinforcement schedule. Again, variable magnitude produced more persistence than continuous magnitude when trials or average reward magnitude were equated. Finally, Experiment 3 compared the effects of chlordiazepoxide, nicotine and haloperidol on persistence with partial (PR) or continuous reinforcement (CR). Chlordiazepoxide decreased persistence with CR and increased persistence with PR; nicotine and haloperidol increased persistence with CR and did not affect persistence with PR. The results suggest that the behavioral similarity of paradoxical behavior in rats and pigeons is a result of convergent evolutionary processes. Persistence in pigeons is discussed in terms of Capaldi's (1967) sequential theory.