9. Behavioral Family Therapy
Behavior therapy may be defined as the clinical application of learning-theory principles to the modification of behavior. London elegantly defines behavior therapy, or “action therapy,” as he classifies it, as “the technical objective of those therapies that manipulate stimulus-response connections in order deliberately to change specific behavior from one pattern of activity to another” (1964, p. 84). Behavioral family therapy, then, may be defined as the application of behavior therapy technology in the clinical intervention to modify maladaptive familial interaction.
John B. Watson (1916, 1924/1970) was largely responsible for formally introducing behaviorism in his lectures at Columbia University in 1912. Watson developed behaviorism from the technology of the natural sciences and from Pavlov’s studies on conditioning, in contraposition to the introspectionist movement in psychology during the first quarter of this century. In Watson’s words:
Behaviorism, as you have already grasped from our preliminary discussion, is, then, a natural science that takes the whole field of human adjustments as its own. Its closest scientific companion is physiology. Indeed you may wonder, as we proceed, whether behaviorism can be differentiated from that science. It is different from physiology only in the grouping of its problems, not in fundamentals or in central viewpoint. Physiology is particularly interested in the functioning parts of the animal—for example, its digestive system, the circulatory system, the nervous system, the excretory systems, the mechanics of neural and muscular response. Behaviorism, on the other hand, while it is intensely interested in all of the functioning of these parts, is intrinsically interested in what the whole animal will do from morning to night and from night to morning.
The interest of the behaviorist in man’s doings is more than the interest of the spectator—he wants to control man’s reactions as physical scientists want to control and manipulate other natural phenomena. It is the business of behavioristic psychology to be able to predict and to control human activity. To do this it must gather scientific data by experimental methods. Only then can the trained behaviorist predict, given the stimulus, what reaction will take place; or, given the reaction, state what the situation or stimulus is that has caused the reaction (1924/1970, p. 11).
The term “behavior,” as used in present-day studies in behaviorism, is usually defined in a broad sense in order “to include a complex of observable and potentially measurable activities including motor, cognitive, and physiological classes of responses” (Bandura, 1969, p. 73).
From the perspective of behavior therapists, behaviors are learned, and the learning of these behaviors obeys psychological principles that have been empirically derived (Bandura, 1969; Bergin & Garfield, 1971; Rimm & Masters, 1974; Ullmann & Krasner, 1969/1975). Examples of these psychological principles include the principles of reinforcement (Premack, 1959, 1965) and the schedules of reinforcement (Ferster & Skinner, 1957; Ayllon & Azrin, 1965, 1968). Thus, a widely supported paradigm in behavior therapy is that “the behaviors traditionally called abnormal are no different, either quantitatively or qualitatively, in their development and maintenance from other behaviors” (Ullmann & Krasner, 1969/1975, p. 2).
The principal assumptions of behavioral learning models, according to Kanfer and Phillips (1970), are: (a) a focus on behavior, that is, on the resultant of the person’s interaction with the environment; (b) the direct intervention of the deviant behavior rather than of the presumed underlying causes of the behavior; (c) all behaviors are subject to the same, empirically derived, psychological principles; (d) the methods of the natural sciences are used in the inquiry into human behavior; (e) observers need to have the ability to make adequate measurements, with no theory-related skills being necessary; and (f) the focus of the intervention is always on the behaviors presently experienced, and not the person’s lived history, or the history of the deviant behavior.
In reviewing Skinner’s operant reinforcement theory in the preceding section we have seen that, basically, learning is produced when an operant response emitted by the organism is reinforced (Skinner, 1969). As a corollary to Skinner’s theory of learning, Bandura (Bandura & Walters, 1963; Bandura, 1968, 1969) has formulated a social learning interpretation of behavior patterns. Bandura’s perspective stresses the role of the environment in the reinforcement of the individual’s responses, thus giving rise to behavioral patterns (1969). Stated in Bandura’s words:
Certain reinforcement practices generate particular behavior, which, due to its aversive properties, in turn creates the very conditions likely to perpetuate it. Thus while nature’s programming ensured that people’s distress would not go unheeded for long, it also provided the basis for the 30 establishment of socially disturbing response patterns. Interpersonal difficulties are most likely to arise under conditions where a person has developed a narrow range of social responses which periodically force reinforcing actions from others through aversive control (e.g., nagging complaints, aggressive behavior, helplessness, sick-role behavior, and emotional expressions of rejection, suffering, and distress, and other modes of responding that command attention). The treatment strategies are quite different depending upon whether one views such behavior in terms of its functional value in controlling the responsiveness of others or as by-products of intrapsychic disturbances. Deleterious reciprocal processes can be best eliminated by withdrawing the reinforcement supporting the deviant behavior and by hastening the development of more constructive means of securing desired reactions from others (1969, pp. 47-48).
Much of the work in behavioral family therapy is based on the above explicated social learning interpretation of Bandura’s, as evidenced by the research reviewed next.
Interventions in behavioral family therapy have been recently directed at parenting (March, Handy, & Hamerlynck, 1976) or directed at the family system (Marsh, Hamerlynck, & Handy, 1976); Berkowitz & Graziano, 1972). Stemming from the first interventions in behavior therapy, the earlier approach has been to focus the efforts of the therapy primarily on the child (Gelfand & Hartmann, 1968; Patterson, 1965; Patterson & Brodsky, 1966; Williams, 1959). Our primary concern in this investigation naturally has been restricted to the field of interventions with families.
The treatment of families presents no special features to the behavior therapist other than an adaptation of the treatment techniques (Patterson, 1969), and a shift in focus from the individual person’s interaction with the environment, to the interaction of a group of people, related by blood and emotional ties, to one another and to the environment. No specific theory is necessary because interactional behaviors are still behaviors, and as such they are subject to the same principles of learning discussed earlier.
Some of the recent research in interventions with families has been reported by Liberman (1970, 1972); Liberman, DeRisi, and King (1973); Wiltz and Eisler (1973); and Eisler and Hersen (1973). To cite an example, Liberman (1972) reported a case study in which he served as consultant to another therapist.
The particular interest in this case is that the target of the intervention was the mother, and not, as typically focused on, the children. The clients were Mrs. G, a 27-year-old woman, her three children, Mr. G, and Mrs. G’s parents. The presenting complaints were Mrs. G’s frequent and severe depressions, accompanied by suicidal thoughts. Also, occasionally this woman was hysterically out of control. Mrs. G had a history of 4 years of conventional psychotherapy and was referred, by her former therapist for hospitalization. A behavioral assessment of the problem was conducted, finding that the family members were positively reinforcing Mrs. G’s maladaptive behaviors.
Thus, the treatment intervention focused on training the family members a contingency management program in which Mrs. G’s complaints were largely ignored; whereas her successful efforts in coping at home were to be verbally reinforced by praise and, more tangibly, with tokens. Mrs. G could exchange her tokens for various privileges, such as time with her therapist in which she could discuss anything, including her complaints. Mr. G’s attentive listening to his wife’s complaints was also to be handled in the same manner. Treatment was successfully terminated after 2 months, although Mr. G continued receiving booster sessions at increasing intervals to maintain the program effectively. Total professional time spent in effecting this treatment intervention was reported to be under 20 hours. Unfortunately, no follow-up data was reported by the author.
Of special interest is the application of behavior techniques in family-oriented crisis intervention by Eisler and Hersen (1973). These authors describe their goals for the intervention to be threefold. The first goal is to develop problem-solving behavior and generate cooperative behavior on the basis of mutual positive reinforcement of adaptive behaviors. The second goal is to train families to effectively communicate their positive and negative feelings concerning the precipitating problem; this is based on the assumption that “effective communication of emotion appears to be very useful in problem solving when it is expressed appropriately and in amounts relative to the eliciting stimulus” (Eisler & Hersen, 1973, p. 112). Finally, the third goal is to help the family members in generalizing the acquired training so that they can use it in other problem situations.
Eisler and Hersen (1973) gave support to their treatment intervention approach by studying the cases of three families. For example, one of their cases involved a middle-class couple, the Joneses, both in their 40s. The precipitating event that led to a crisis situation seems to have been Mr. Jones’s 16-year-old son Tom, from a previous marriage, who came to live with the family. Two years later, Mrs. Jones was exhibiting signs of nervousness, depression, crying spells, and hysterical outbursts. This woman had been given other types of psychotherapy, including psychotropic medication and electroconvulsive treatment, but with no significant improvements reported. At the time of the intervention, Mrs. Jones was threatening to file for a divorce unless Tom left the home.
In the treatment approach advocated by Eisler and Hersen (1973), the family was asked, during the first session, to re-enact a typical situation that gave rise to family conflict. This role-playing was videotaped for future feedback to the family members and served as a sample of the deviant interaction patterns occurring within the family.
This sample of behavior provided the therapists with the necessary information (baseline) on the behavioral patterns, and at the same time provided the basis for treatment. In other words, the family members were specifically shown their deviant behaviors. Later, the therapists used a modeling procedure designed to provide training in problem-solving behavior and cooperative behavior. The family members were then offered contingency management procedures to use in dealing with some of their specific deviant behaviors. Treatment lasted for 10 sessions.
A follow-up 3 months later indicated that the family members had maintained their recently acquired problem-solving behaviors. The authors argued that treatment of a situation of familial crisis requires a rapid assessment of the behavioral difficulties, and that this could be accomplished by the in vivo enactment, or role playing, of the family members’ interaction during an issue of conflict.
In a better controlled study, Alexander and Parsons (1973) used a short-term behavioral intervention approach with delinquent families. Their treatment intervention 35 consisted of an analysis of the behaviors that maintained the deviant behaviors and a modification procedure designed to strengthen clarity and precision in the family communication patterns, as well as increased reciprocity and problem-solving behaviors. The modification of behaviors was effected with a program of contingency management to reinforce adaptive behaviors and to decelerate deviant ones.
A total of 99 families were referred by a juvenile court for offenses such as running away, habitual truancy, shoplifting, or possession of soft drugs, alcohol, or tobacco. Of the total number of families, 46 were randomly assigned to the experimental group; in addition, 30 families were assigned to one of three comparison groups: client-centered family program; psychodynamic family program; and a no-treatment control group.
Alexander and Parsons (1973) used both process and outcome measures. The process measures were based on the performance of three tasks assigned to the family members, during which their interaction was recorded with individualized voice-actuated microphones. The authors hypothesized that the experimental treatment families would exhibit a lower average within family variance of talk time across groups, a decreased amount of silent periods, and, finally, a greater frequency of interruptions, as mechanically measured by the recording apparatus. The outcome measures included the examination of juvenile court records for recidivism.
The authors reported finding significant differences for each of the three process variables with families who experienced the behavioral intervention, which demonstrated significantly lower variance, that is, equality, in talk time, less silence, and more interruptions. The outcome measures were conducted 6 to 18 months after treatment termination and measured rates of recidivism. The no-treatment controls demonstrated a 50% rate of recidivism; the client-centered family program showed a 47% rate of recidivism; and the psychodynamic family program showed a 73% rate of recidivism. The short-term behavioral group demonstrated a significant reduction to 26% rate of recidivism. These figures were compared to the countywide 51% rate of recidivism and the 48% rate demonstrated by a post-hoc, selected, no-treatment control group. Alexander and Parsons concurred on the comparative efficacy of their methods over the other treatment modalities discussed in their study.
A most productive and consistent program of research in family therapy has been carried out by Patterson and his collaborators (Patterson, 1971, 1975, 1976; Patterson, Shaw, & Ebner, 1969; Patterson & Cobb, 1971; Patterson, Ray, & Shaw, 1968; Patterson & Reid, 1970, 1973; Patterson, Reid, Jones, & Conger, 1975; Patterson, Cobb, & Ray, 1969, 37 1972; Arnold, Levine, & Patterson, 1975). Most of Patterson’s work is based on one of the main assumptions of the social learning approach, namely, “the social environment provides positive social reinforcers contingent upon deviant child behaviors that are sufficient to maintain these behaviors” (Patterson, 1971, p. 752).
Patterson, Reid, Jones, & Conger (1975), in an effort to communicate procedures more explicitly, have published a manual descriptive of their social learning approach to family intervention. For example, an earlier version of their treatment approach is described as including the following procedures:
…an intake evaluation; a discussion of the referral problems; two weeks of baseline or pretreatment observations of the child and family in the home by experienced observers; and the development of a family interaction code for collecting these observation data.
The observation data provided information on the frequency or base rates of coercive child behaviors against which the effects of treatment could be compared. After these baseline observations were collected, the parents were given a programmed set of materials to study which described the child management procedures. The parents were then taught, in the laboratory, to carefully define, track, and record targeted deviant and prosocial behaviors. Parents were monitored daily by telephone calls to insure their participation in this and other phases of treatment. The parents were then taught, at the office, to design and carry out modification programs in the homes. For most of the families, it was thought necessary to actually go into the home to role play or model the procedures directly for the parent (Patterson, Reid, Jones, & Conger, 1975, p. 10).
In a more advanced state in the development of their procedural system, Patterson, Cobb, and Ray (1973) conducted a study with five families, each of which included a child who was excessively aggressive, asocial, or out of control. The baseline procedure consisted of 10 observations during a 2- to 3-week period prior to the onset of treatment. Two more observations were made at the conclusion of treatment, and then at monthly intervals for 6 months.
The outcome measure was based on an observational code-system of 14 deviant behaviors on which the children had been observed and rated during the baseline period, as well as during the follow-up sessions. The sum of the 14 behavior codes yielded an over-all rate of behavioral deviancy per minute. The findings reported, based on four of the five families, indicated that there was a 60 to 75% reduction from the rates of deviant behavior obtained with the baseline measure. In addition, three of the four families maintained the therapeutic gains over the 6-month follow-up period. The total average of therapist’s time per family was reported to be 22.8 hours to produce the effects noted above.
In other recent and well-controlled investigations, Patterson (1976) reported his findings in treating families with coercive or aggressive children. In addition, a replication study with 10 families (Patterson & Reid, 1973) was conducted to verify the earlier findings of Patterson, Cobb, and Ray (1973). Because of the inherent similarity of these two studies (Patterson, 1976; and Patterson & Reid, 1973), they are reported here concomitantly.
In addition to the treatment approach procedures used by Patterson, Cobb, and Ray (1973), Patterson (1976), and Patterson and Reid (1973) used a programmed training manual (Patterson, 1971/1975) which they gave the parents to study and be tested on, as a prior requisite to treatment. Also some of the families in these studies required one home visit during the course of treatment. The total number of hours of professional time spent on these two studies were reported at 25.7 hours and 31.4 hours, respectively. These hour totals of professional involvement did not differ significantly from the total of 22.8 hours reported by Patterson, Cobb, and Ray (1973). The outcome criteria measures in these two studies were essentially similar to the measures used in the report by Patterson and Reid (1973). The findings in both studies demonstrated that approximately 2/3 to 3/4 of the children showed a major reduction in their frequency of exhibited deviant behaviors.
Patterson and his collaborators posit that, over-all, their research findings appear to indicate that their social learning family therapy approach produces: (a) a 40 decrease in the frequency of deviant behaviors exhibited by the child, and this is maintained at follow-up time; (b) a modest reduction in the frequency of deviant behaviors exhibited by the family members; (c) a more positive parental perception of their problem children; (d) less reinforcing parental consequences for the child’s deviant behavior; (e) a more effective parental use of aversive consequences for the child’s deviant behavior; (f) a more active role of fathers in controlling their children’s deviant responses; and (g) a posttreatment indication by the majority of the mothers that the family was “happier,” as well as an indication by all the mothers that their children were “improved” (Patterson & Reid, 1973; Patterson, Reid, Jones, & Conger, 1975).
© 1976 Angel Enrique Pacheco, Ph.D., C.Psych. All Rights Reserved.
We are here to help you…
Learn to Live Better ®