Aversive schedules with independent probabilities of reinforcement for responding and not responding by rhesus monkeys: II. Without signal.

Abstract

The probabilities of aversive stimulus presentation given the occurrence or the nonocurrence of a lever-press response were manipulated independently. These manipulations generated a variety of schedules of uncued aversive stimulation. Ten rhesus monkeys were divided into 5 groups, each group receiving a different sequence of the 2 probabilities and thereby a different sequence of electric shock presentation schedules ranging between avoidance and punishment. The schedules generated in this manner provided systematic control both of the amount of response facilitation that occurred when response-produced shocks were first introduced following avoidance training and of the postfacilitation response rate decline. Additionally, when the probability of shock presentation following a response was less than that for not responding, scalloped response patterns occurred; when the relative shock probabilities were reversed, bipeak response patterns were observed.

Model of response-reinforcer contingency.

“…The present experiment demonstrates that the performance of subjects in two groups with the same values of u and v is related to the probability of a criterion response (w), which is controlled by a percentile reinforcement schedule. Previous operant studies using the Rescorla index (u -v; e.g., Hammond, 1980;Kadden, et al, 1974;Kop, 1974;Neffinger & Gibbon, 1975) cannot provide good experimental evidence for the importance of a response-reinforcer contingency variable. Even if the variation in the probability of a response was fortuitously small, so that resulting phi values were in the same rank order as the (u -v) values, "one cannot.…”

Aversive schedules with independent probabilities of reinforcement for responding and not responding by rhesus monkeys: III. Recovery of avoidance baseline.

“…[Mean rates include both t° and t* periods. The coordinates at the top show the p(S~B) combinations that were tested in a previous experiment (Kadden et al, 1974).] EXPEKIMENT 2: RECOVERY OF AVOIDANCE RESPONDING FOLLOWING UNSIGNALED S~B SCHEDULES This experiment was a continuation of the study by Kadden et al (1974).…”

“…The coordinates at the top show the p(S~B) combinations that were tested in a previous experiment (Kadden et al, 1974).] EXPEKIMENT 2: RECOVERY OF AVOIDANCE RESPONDING FOLLOWING UNSIGNALED S~B SCHEDULES This experiment was a continuation of the study by Kadden et al (1974). As in Experiment 1, each session was subdivided into T cycles containing i A and P portions.…”

“…Method Subjects and apparatus. The subjects were the same 10 rhesus monkeys that served in the Kadden et al (1974) study. They were each placed, unrestrained, in primate test chambers for experimental sessions.…”

“…Two previous experimental reports on the use of electric shock as the aversive reinforcer, S~R, have described the effects of independently varying the probability of shock delivery, p(S~R) for responding (R) or not responding (R). One of those studies employed an exteroceptive signal preceding S~R; the other did not (Kadden, Schoenfeld, & Snapper, 1974;Kop, 1974). Control of both response rate and response patterning was shown to be dependent upon the 2 p(S~R) values and upon the sequence of these conditional probabilities as presented to individual subjects.…”