Some effects on human behavior of aversive events

Abstract

Two studies examined aggressive behavior in humans. In Expt 1, 10 adults working on a plunger pulling task could receive a 3.5 mA shock at 75% probability every two min. The shock was unrelated to their plunger pulling behavior. Subjects could press a toggle switch to deliver electric shock to the experimenter, who was visible to the subjects as an alleged observer. Three sessions in which no shock was delivered alternated with two sessions in which shock was delivered. In shock sessions subjects pressed the toggle switch and shocked the observer at an average rate more than nine times higher than in the non-shock sessions. In Expt 2, baselines of aggressive behavior were collected for three consecutive 120 or 90 min segments daily. The subjects were two children, and a single-subject repeated measures multiple baseline design was used. After approximately two weeks of baseline, a brief time-out was made contingent upon aggressive behavior which occurred in the first segment each day, and non-contingent time-outs roughly yoked to the time-outs in the first segment were delivered in the third segment each day. Other conditions included a return to baseline in all three time segments, responsecontingent time-out in the first segment with baselin conditions in the other two segments, response-contingent time-out in the first segment and the first hour of the second segment, with noncontingent time-out in the second hour of the second segment and in the third segment, and responsecontingent time-out in all three segments. Non-contingent time-out seemed to control an above baseline rate of aggressive behavior in the segments in which it was programmed, as well as in the segment adjacent to that in which it was programmed, and in which time-out was not programmed at all. Results were interpreted as replicating the animal laboratory findings relating non-contingent aversive stimulation to aggressive behavior. The specific controlling relationship in the latter case needs further analysis to be interpretable.