Partial reinforcement and autoshaping of the pigeon's key-peck behavior

Abstract

Pigeons were exposed to continuous and partial reinforcement (25%) schedules in an autoshaping procedure. The number of trials and/or pairings per session were equated across two factorial group sets and all groups were trained for a fixed total number of acquisition trials (192). When acquisition was analyzed as a function of reinforcements delivered, the groups receiving fewer number of trials per session acquired the response faster, independently of the percentage of reinforced trials. Moreover, the number of reinforcements to satisfy an acquisition criterion was a monotonic function of the number of trials per session but not of the schedule. The results are discussed in relation to the neocontinuity and the scalar expectancy theories.