VARYING REINFORCEMENT MAGNITUDE ON INTERVAL SCHEDULES

In Experiment 1, rats were trained on either a random-interval or a va riable-interval 60-sec schedule of reinforcement, and reinforcement ma gnitude was varied across conditions between one and four pellets. Alt hough the two schedules maintained different patterns of behaviour, pa tterns and rates of responding were not systematically affected by the variation in reinforcement magnitude. In Experiment 2, a regulated pr obability interval schedule that generated similar rates of reinforcem ent to those of the schedules of Experiment 1 was used, with the patte rn of behaviour generated resembling that typical of a random-interval schedule. Changing reinforcement magnitude again produced few systema tic changes in behaviour. In Experiment 3, a variable-ratio schedule w as used within a procedure that otherwise resembled that of Experiment s 1 and 2. Increasing the reinforcement magnitude now decreased the ra tes of responding, and examination of the patterns of responding showe d that this came about because rates of responding were higher early i n the interreinforcer interval in the one-pellet condition. These expe riments demonstrate the insensitivity of behaviour under interval sche dules to changes in reinforcement magnitude and suggest the operation of mechanisms different from those engaged by ratio schedules and disc rete trial learning procedures.