Free operant avoidance behavior in hooded rats: IRTs and response chains

Abstract

Twenty-five hooded rats were given 50 avoidance training sessions with leverpress IRTs ob­ tained during as well as between sessions. Five qualitatively different shapes to the IRT pro­ files were found: single- or double-peaked shapes, positively or negatively accelerated shapes, and U-shaped distributions. Both the specific shape of the profile and the stability of that shape during a session were related to the avoidance proficiency of an individual rat. Six rats received two additional training sessions, with the final 30 min of each videotaped to obtain frequency counts of nine behavior categories. The observational data showed that avoidance-proficient rats typically incorporated the leverpress requirement into repetitive response chains. Under the free operant avoidance procedure (Sidman, 1953), each criterion response postpones an electric shock for a fixed period (the R-S interval). If the re­ sponse is not emitted, shock is given periodically (the s-s interval). This schedule permits the subject to develop a rate of response that leads to relatively low frequencies of unavoided shock. Considerable attention has been given to the analysis of the behavior that develops under this avoidance schedule, and recently the examination of changes in avoidance behavior that might occur within a training session has been of particular interest. For example, after the warm-up period (a repeated, temporary period of relatively poor avoidance at the outset of each session), changes in leverpress avoidance behavior correspond to the avoidance proficiency of the individual rat. Roberts (1978) used shock rate values to designate individual rats as good avoiders (low shock rates), nonavoiders (high shock rates), or poor avoiders (a broad range of rate values in between). The response and shock rates of individual nonavoiders and good avoiders were relatively stable over the post-warm-up period, but poor avoiders showed progressive improve. ments in these measures during the session, but not be­ tween sessions. The picture of post-warm-up avoidance behavior may be somewhat different when using a different index of avoidance, the distribution of leverpress responses within the R-S interval (interresponse times: IRTs).