THE RELATION BETWEEN VARIABLE-INTERVAL SCHEDULE VALUE AND SUBSEQUENT PROGRESSIVE-RATIO PERSISTENCE

The effect of variable-interval reinforcement rate on subsequent progr essive-ratio responding was investigated by exposing five groups of ra ts to 15-session blocks of variable-interval training followed by a si ngle progressive-ratio ''test'' session. Progressive-ratio test sessio ns began with a response requirement of 25 and incremented by 25 follo wing each reinforcer. Sessions terminated when no responding occurred for 15 min. The reinforcement rate arranged by the variable-interval s chedule was changed following exposure to each block of training and t est sessions according to a Latin square design. Each subject received two blocks of training with variable-interval schedules arranging rei nforcement rates of .25, .5, 1.0, 2.0, and 4.0 reinforcers per min. Re sponse rates during the final five sessions of each variable-interval exposure was an increasing function of variable-interval reinforcement rate. Similarly, total responses emitted and response rate during pro gressive-ratio test sessions increased as a function of variable-inter val reinforcement rate. In contrast, the amount of time required to sa tisfy the 15-min extinction criterion used to terminate test sessions decreased as variable-interval reinforcement rate increased. Similarly , the ratio of response rate during test sessions to the response rate during the five variable-interval sessions preceding the test was inv ersely related to the reinforcement rate arranged during the variable- interval sessions. Responding during test sessions persisted longer an d response rate declined less following exposure to the lowest reinfor cement rate variable-interval schedules than following the richer para meters. The use of occasional test sessions interpolated between block s of training sessions is a promising method for assessing the relatio n among schedule parameters and performance under subsequent condition s. (C) 1994 Academic Press, Inc.