Extinction, DRO, and response—cost procedures for eliminating self-injurious behavior: a case study

Abstract

In recent years, self-injurious behavior in retarded and emotionally disturbed children has received a great deal of attention from behaviorally oriented psychologists and educators. The use of methods developed for treating these types of behavior, such as electric shock and other forms of physical punishment, have recently been strongly challenged by parents, guardians, and other child advocates. As a result, many of these methods are no longer available to the behavioral psychologist even though the incidence of self-injurious behavior is common among retarded and emotionally disturbed children. Thus, a need has arisen for new approaches to eliminating self-injurious behavior that would be both humanistically acceptable and educationally practical. From an operant conditioning point of view, self-injurious behavior is considered to be partly a function of its social consequences. That is, the rates of self-injurious behavior can be altered through variations of the results of the behavior (Lovaas et al., 1965). A serious problem that has led us into our present dilemma of having few acceptable treatments is that while positive reinforcement will increase self-injurious behavior, the withdrawal of that reinforcement will not always eliminate or even reduce it (Risely, 1968). Since the removal of positive reinforcers does not consistently reduce self-injurious behavior, researchers have resorted to using punishment such as electric shock (Bucher and Lovaas, 1968; Tate and Baroff, 1966). These aversive treatments, as is well known, have been found to be consistently effective in removing even the most chronic and intractable forms of self-injurious behavior. Today, however, electric shock or some of the other equally aversive therapies are reserved, even in the more experimentally permissive institutions, for those individuals whose self-injurious behavior is threatening to their life. In spite of the limitations on the use of highly aversive conditioning procedures, very little research has been reported on less aversive forms of conditioning procedures for eliminating self-injurious behavior. Several reasons account for this paucity of research. First, aversive conditioning is against the philosophy of most educators and its use is generally banned either by institutions, school systems, boards of education, or grant specifications. Second, the mild-to-moderate forms of self-injurious behavior, for which highly aversive conditioning appears to be a ‘punishment’ outweighing the ‘crime’, are typically not alarming and usually do not interfere too greatly with the training and education of the student. Third, and most important, most teachers and psychologists have no definitive programs for humanely dealing with behavior that in the past has been responsive only to highly aversive treatments. An important need, then, is to develop several workable approaches for eliminating mild-to-moderate self-injurious behavior which can be applied by a trainer or teacher and are compatible with school or institutional bans against aversive conditioning. In response to this need, the following experimental case study was conducted to compare the efficacy of three possible operant approaches for eliminating self-injurious behavior in the trainable mentally retarded. Of the three methods compared the response—cost procedure discussed by Azrin and Holz (1966) was of special interest because of their contention that it was analogous in effectiveness to electric shock. The purpose of the present research was to determine whether there are significant differences in effectiveness among extinction, differential reinforcement of other behavior (DRO), and response—cost procedures.