Intermittent reinforcement of a continuous response.

Six rats were trained with food deliveries contingent upon their pressing a lever and holding it down for either fixed or variable cumulative durations. Fixed-hold requirements ranged from 15 s to 90 s over experimental conditions; variable-hold requirements ranged from 15 s to 120 s. At most long and intermediate values, variable-hold requirements maintained more lever holding than fixed requirements. At the longest hold requirements studied, more lever holding was maintained by variable requirements than by fixed requirements of equivalent mean length for each rat. Postreinforcement-pause duration increased with lever-holding time for both fixed- and variable-hold requirements. At comparable lever-holding times per reinforcer, longer pauses typically were produced by fixed requirements than by variable requirements. Data from this study on the maintenance of responding, temporal response patterns, and postreinforcement pausing are comparable to those obtained with intermittent reinforcement of discrete responses. These findings suggest that the response-reinforcer relation specified by a reinforcement schedule is a fundamental determinant of responding, whether responding consists of discrete units or of continuous activity.