A COMPARATIVE STUDY EVALUATING THE DIFFERENTIAL EFFECTIVENESS 0F BEHAVIOR MODIFICATION TREATMENT GROUPS AND GROUPS WITH AN ADDED COMPONENT, RATIONAL EMOTIVE THERAPY Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY MARY ALICE COLLINS 1974 vw~ q This is to certify that the thesis entitled A GMPARATIVE STUDY EVAUIATEK; THE DIFFERENTIAL EFFECTIVENESS OF BEHAVIOR [VDDIFICATIGNI TREATMENT GROUPS AND GROUPS WITH AN ADDED C(MPONENI', RATIONAL ENUl‘IVE THERAPY presented by Mary Alice Collins has been accepted towards fulfillment of the requirements for Ph.D. degreein Social Science me .7. were” Major professor Date W 0-7 639 ABSTRACT A COMPARATIVE STUDY EVALUATING THE DIFFERENTIAL EFFECTIVENESS OF BEHAVIOR MODIFICATION TREATMENT GROUPS AND GROUPS WITH AN ADDED COMPONENT, RATIONAL EMOTIVE THERAPY BY Mary Alice Collins The problem of interest in this research was to determine whether adding cognitive restructuring to behavior modification outpatient treatment groups increases behavioral change. The cognitive restructuring selected was Rational Emotive Therapy and the model of behavioral social group work used was the one develOped by Lawrence and Sundel (1972) and further refined by Lawrence (1973). Thirty-six mental health outpatient clients were randomly selected and randomly assigned to one of three treatment groups. Fourteen clients actually completed treatment. The first treatment condition was behavior modi- fication plus Rational Emotive Therapy, the second behavior modification only, and the third behavior modification plus Rational Emotive Therapy excluding pretests. Clients were measured on pregroup and postgroup baselines of two beha- viors, a self-report questionnaire, a series of three vignettes, the Rational Behavior Inventory developed by fl! . a? Mary Alice Collins .9 {V'flh h- I i: 3' f U i- V. II L. Shorkey and Whiteman (1973), the Collins-Curran scale of rational thinking, and the Curran therapist evaluation form. Clients attended six 2%-hour treatment sessions. Each session was planned in advance according to a prescribed treatment regimen. The first aspect of the study was concerned with determining whether or not behavioral change was increased in the combined treatment condition. As predicted, beha- vioral changes were significantly higher in the experimental group, although change occurred in the desired direction in both treatment conditions. The third group lost significance because of a loss in membership caused by external circum- stances. The second hypothesis in the study predicted that rational thinking would be increased in the Rational Emotive Therapy plus behavior modification condition over the beha— vior modification only condition. No significant results were found. The rational Behavior Inventory showed an increase in rational thinking in the experimental group from pretest to posttest, but the difference was not significant. The final aspect of the study tested for client generalization of behavioral problem—solving methods. Results indicated that in two instances generalization did not occur. In the third case, where generalization did occur, the problematic situation was broader in scope than in the first two instances. A COMPARATIVE STUDY EVALUATING THE DIFFERENTIAL EFFECTIVENESS OF BEHAVIOR MODIFICATION TREATMENT GROUPS AND GROUPS WITH AN ADDED COMPONENT, RATIONAL EMOTIVE THERAPY BY Mary Alice Collins A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Social Science 1974 ACKNOWLEDGMENTS Appreciation is extended to Dr. Victor L. Whiteman, chairman of the committee, and Dr. Clayton Shorkey, who served as chairman for two years before he was appointed to the faculty of the University of Texas. The writer also acknowledges the contributions of Dr. Donald Olmsted and Dr. Mark Rilling, who served on the committee. ii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . Chapter I. INTRODUCTION . . . . . . . . . . . . . . . . II. A HISTORICAL PERSPECTIVE OF SOCIAL GROUP WORK . . . . . . . . . . . . . . . . III. RELEVANT LITERATURE . . . . . . . . . . . . Behavior Modification . . . . . . . . . . Group Rational Emotive Therapy Literature IV. METHOD . . . . . . . . . . . . . . . . . . . Hypotheses . . . . . . . . . . . . . . . . Operational Definitions . . . . . . . . . Selection of Subjects . . . . . . . . . . Independent Variables . . . . . . . . . Dependent Variables . . . . . . . . . . Design . . . . . . . . . . . . . . . . . Procedure . . . . . . . . . . . . . . . Initial Contact . . . . . . . . . . . . Telephone Contact . . . . . . . . . . Initial Interview . . . . . . . . . . . Initial Interview Checklist . . . . . . Second Interview . . . . . . . . The Six Group Sessions . . . . . . . . Final Interview . . . . . . . . . . . V. RESULTS . . . . . . . . . . . . . . . . . Baseline Behaviors . . . . . . . . . . . . Analyses of Variance of Baseline Behaviors Analyses of Variance of Measures of Rational Thinking . . . . . . . . . . . iii Page vii 10 10 l4 l6 l6 17 20 22 22 24 24 24 25 25 25 27 28 29 30 3O 42 43 Chapter Analyses of Variance for Generalization of Behavioral Problem Solving . . Analyses of Variance of Client Reports and Therapist's Estimate of Improvement Correlations of Dependent Variables Validation of Therapist Interventions Summary of Results . . . . . . . . VI. DISCUSSION . . . . . . . . . . . . . Research in Clinical Settings . . . Discussion of and Implications of the Results . . . . . . . . . . . . Some Additional Effects of Client Participation in the Research . . Suggestions for Future Research . . Conclusion . . . . . . . . . . . . . APPENDICES . . . . . . . . . . . . . . . . . . A. QUESTIONNAIRES . . . . . . . . . . . B. ANALYSES OF VARIANCE TABLES . . . . . C. INDIVIDUAL SCORES . . . . . . . . . . . BIBLIOGRAPHY . . . . . . . . . . . . . . . . . iv Page 47 51 54 54 55 57 57 61 66 67 70 71 72 88 92 97 LIST OF TABLES Table Page 1. Analyses of Variance of Baseline Behaviors . . 43 2. Analyses of Variance for the Rational Behavior Inventory . . . . . . . . . . . . . . 44 3. Posttest Analyses of Variance Between Groups on the Rational Behavior Inventory . . . . . . 45 4. Comparison on the Collins-Curran Scale-- Analyses of Variance . . . . . . . . . . . . . 46 5. Posttest Comparisons Between Groups on the Collins-Curran Scale . . . . . . . . . . . . . 46 6. Analyses of Variance of Pre- and Posttest Comparisons on Generalization Vignettes . . . 48 7. Analyses of Variance Posttest Group Comparisons on Generalization Vignette l . . . 49 8. Analyses of Variance Posttest Group Comparisons on Generalization Vignette 2 . . . 49 9. Analyses of Variance of Posttest Group Comparisons on Generalization Vignette 3 . . . 50 10. Mean Scores on Client Problem Checklist . . . . 51 ll. Analyses of Variance of Clients' Problem Checklists . . . . . . . . . . . . . . . . . . 52 12. Curran Therapist Rating Scale Analyses of Variance . . . . . . . . . . . . . . . . . . . 53 13. Correlations . . . . . . . . . . . . . . . . . 54 Appendix Tables A. Analyses of Variance of Number of Client Relationships . . . . . . . . . . . . . . . . 89 Table Page B. Analyses of Variance on Clients Wishing Help at the Present Time . . . . . . . . . . . 89 C. Analyses of Variance on Source of Help . . . . 90 D. Goal Attainment Scale . . . . . . . . . . . . . 91 vi 10. ll. 12. 13. 14. 15. 16. 17. 18. LIST OF FIGURES Page Number of Periods of Anxiety Per Week . . . . . 31 Number of Angry Outbursts Per Week . . . . . . 31 Number of Times Per Week Client Presented Himself as a Loser . . . . . . . . . . . . . 32 Number of Times Per Week Behavior Criticism Used . . . . . . . . . . . . . . . . . . . . 32 Number of Periods of Depression Per Week . . . 33 Number of Times Per Week Client Was Assertive . 33 Number of Times Per Week Client Was Assertive . 34 Number of Times Per Week Client Was Assertive to Boyfriend . . . . . . . . . . . . . . . . 35 Number of Dependent Actions Per Week . . . . . 35 Number of Uninterrupted One-Half Hour Study Periods Per Week . . . . . . . . . . . . . . 35 Number of Assertive Behaviors Per Week . . . . 36 Number of Snacks Per Week . . . . . . . . . . . 37 Number of Approach Responses to Women Per Week . . . . . . . . . . . . . . . . . . 38 Number of Anxiety Attacks Per Week . . . . . . 39 Number of Dishwashing Arguments Per Week . . . 39 Number of Depression Periods Per Week . . . . . 39 Number of Times Per Week Client Stayed in Bed All Day . . . . . . . . . . . . . . . . . 40 Number of Times Per Week Client Was Verbally Assertive . . . . . . . . . . . . . 4O vii Figure 19. 20. 21. 22. Number of Times Per Week Client Avoided People . . . . . . . . . . . . . . . . . Number of Conversations Initiated Per Week . . . . . . . . . . . . . . . . Number of Anxiety Attacks Per Week . . . . Number of Periods of Rigid Behavior Per Week viii Page 41 41 42 42 CHAPTER I INTRODUCTION Behavior modification has become one of the major frameworks for social group work practice. Lawrence and Sundel (1972) applied behavior modification principles to social group work with adults. Rose (1967) provided a behavioral group model for working with children. The par- ticular problem of interest in this research is to determine whether adding a cognitive element to the model of social group work developed by Lawrence and Sundel (1972) and further refined by Lawrence (1973) increases behavioral change. Despite increasing interest among behaviorists in cognitive restructuring (Lazarus and Meichenbaum, 1971), this interest has not yet been strongly reflected in the behavioral social group work literature. The specific cognitive component selected for this study is Rational Emotive Therapy developed primarily by Albert Ellis (Ellis and Harper, 1961). The intent of this research is to ascertain whether adding a cognitive dimension to therapeutic group input actually increases behavioral change in clients. The population of interest in the research is the typical outpatient client in a mental health center or family agency. The aim of the research is to test out behavioral l and cognitive treatment in an actual clinical setting, with all of the resultant hazards that exist away from either an inpatient unit or a laboratory setting. The research for the project was conducted at Ingham Medical Mental Health Center, which serves a client pOpulation over 18 years of age. Actual clients were used as subjects in an attempt to increase the generalizability of the research. Clients at a community mental health center or a family agency differ from university students, often used in laboratory studies, in a number of significant ways. These differences include educational level, socio- economic status, degree of verbal ability, age, life style, and value systems. L'Abate (1969) spoke of the "behavior without its context" (L'Abate, 1969, p. 482). He went on to argue for taking client characteristics into context in evaluating behavioral techniques. The present research is an extension of the experimental method to the clinical setting. Using clients as subjects creates some problems, however. One problem is that there is a limited client pop- ulation at any one mental health center or family agency. This becomes particularly problematic when the target of research is groups. Having a large number of groups for comparison becomes nearly impossible unless the researcher uses several settings in one experiment. Using several settings is hampered by financial considerations. A second problem is that clients come to an agency primarily to solve their problems, not to aid the researcher or to be suitable subjects according to research standards. An example of this type of problem is that subjects may drop out before the research is completed because they have achieved their objectives. There is an essential difference between client status and subject status. The core differ- ence is that the client has a real problem, for which he has sought help. I A third problem with clinical research is that one cannot keep outside factors from influencing outcome. For example, a strike at Oldsmobile would be an influential event in the lives of many clients, over which a researcher would have no control. Perhaps this category of research difficulty is the reason why a number of studies are con— ducted in closed settings, such as state and Veterans' Admin- istration hospitals. However, the question can be raised: What is the difference between those people in closed insti- tutions and those in Open settings? Despite the difficulties of doing research in out- patient settings, more such research is needed. Behavior modification and cognitive therapy need to be examined in the context of community treatment. Research in a clinical setting can provide valuable information to social group workers. For example, will clients pay for the type of group conducted? Also, will the therapist be able to adhere to the prescribed treatment regimen? The present research provides an experimental design within a community context aimed at measuring the effect of adding Rational Emotive Therapy to behavioral social group work. CHAPTER II A HISTORICAL PERSPECTIVE OF SOCIAL GROUP WORK Social group work has undergone several major trans- formations since its beginning between 1900 and 1910. The purpose of early group work was social reform, and most early group workers practiced in settlement houses (Briar, 1972). Examples of early leaders in group work include Canon Barnett and Jane Adams, who were also leaders in the settlement house movement (Wilson and Ryland, 1949). Canon Barnett established the first settlement house, where working men and university men could come together, share ideas, and work for common economic and social goals (Wilson and Ryland, 1949). Early group work emphasized a situational view of problematic behavior. Wilson (1956) cogently summarized the time at which group work emerged: Social Group Work, as one of the methods of the social work profession, was introduced during the first quarter of this century. It emerged at a time when there was a renewed dichotomy within the profession between those social workers who primarily regarded the causes of social problems as within peOple, in contrast to others who located the causes primarily within the social sit— uation in which peOple with problems were living (p. 27). The aim of early group work was to gather together peOple with similar problems, both to achieve personal growth and to act on the environment. Following World War I there was a conservative atmos- phere in the United States which was reflected in group work practice. Settlement houses became less the locus of group work activity and social reform less important as a purpose. The shift was to leisure time agencies, such as the Y.M.C.A., Y.W.C.A., Boy Scouts, and Girl Scouts. Social participation, democratic processes, personal growth, and direct contact among people of varied backgrounds were emphasized to the same extent. A key figure in the development of group work as a method of social work was Grace Coyle. Coyle taught at Western Reserve, which offered the first courses in group work in 1923 (KonOpka, 1961). Prior to 1946, group work was not clearly aligned with the profession of social work. Group work also had roots in recreation and education. In 1946 Coyle addressed the members of the American Association for the Study of Group Work in Buffalo, New York, on the topic, "On Becoming Professional" (Coyle, 1947). In this speech she argued that there was a common factor between casework, group work, and community organization, and that factor was that all three are based on understanding human relations. During the same conference the American Association for the Study of Group Work voted to affiliate withsxxfijfl.work. Other leaders in the World War II and post-World War II eras were Harleigh B. Trecker, who published Social Group Work in 1949; Gisela Konopka, who wrote Therapeutic Group Work with Children in 1949; and Gertrude Wilson and Gladys Ryland, who wrote Social GroupiWork Practice, also in 1949. It is interesting to note KonOpka's view of the World War II era and its influence on social group work. She stressed the importance of individualization in the group and contrasted it to her experiences in Nazi Germany. She stated: . . . I must say that my first encounter with social group work in 1941 was a revelation. Having just come from a society that seemed to present an inescapable gulf between the individual and the group--which insisted that the individual be sacrificed to the interest of the group-—I found the concept of individ- ualization in and through the group exhilarating (Konopka, 1961, p. 9). This focus on individuals in the group rather than the group solely as a whole has been prevalent in the his- tory of group work. The authors cited thus far developed a body of knowledge that heavily emphasized personal growth and social contribution. Coyle stated: We believe that each individual should be encouraged to develop his own powers to the fullest and we believe he should freely devote those powers to the social good by full participation in the society in which he lives (Coyle, 1947, p. 67). It is interesting that early group work was not par- ticularly identified with a labeled population. Group members were considered citizens rather than clients. Later on, when the rehabilitative approach was accepted, the term client was most often used. Perhaps what was a gain for the profession was a loss for clients, in that a possible stigma exists with the term client. The fifties saw a struggle between the proponents of therapeutic group work, such as Robert D. Vinter, and the prOponents of a more traditional approach as practiced in the leisure time agencies (Briar, 1972). Those with a more traditional view often characterized group work as being for the healthy and casework as dealing with the client pOpulation with problems. According to Papell and Rothman (1966), there are three models of social group work: the social goals model, the remedial model, and the recipro- cal model. Thus, even in the sixties and seventies, three elements are preserved in different forms of group work: social action, individual growth, and problem solving. Individual problem solving was the last to be introduced into group work. The Vinter model, which is consistent with both the behavioral and cognitive aspects used in this research, emphasizes: (l) the individual as the focus of change, (2) specificity of goals, (3) contract, (4) the group as a means to change, (5) stages of group develOpment, and (6) interventions in the social environment (Vinter, 1967). Since the development of the Vinter model, the application of behavior modification to social group work has occurred. This writer has noted that advocates of the Vinter model and advocates of behavioral social group work are careful not to see the two approaches as mutually exclu- sive. One could predict an article in the future by someone presenting a blend of the two approaches as a unified approach. CHAPTER III RELEVANT LITERATURE Behavior Modification In contrast to much of both the current and histori- cal social group work literature, more than 70 journal articles, mainly by psychologists, have been published since 1960 on the application of behavior modification to groups. Generally, behavior modification techniques are as effective when applied in a group as when applied individually. More than half of the journal articles dealt with operant tech- niques, which are the concern of this research. A number of studies have been conducted to demon- strate the effectiveness of positive reinforcement in a group. Abudabbeh, Prandoni, and Jensen (1972) worked with a group of five adolescent boys and demonstrated that chips that could be exchanged for money, candy, or telephone calls could be used to increase different units of verbalization; specifically, statements related to self, feelings, personal problems, and group interactions. Aiken (1965) and Bachrach, Candland, and Gibson (1961) in separate laboratory studies determined that the frequency of verbalizations can be increased through positive reinforcement. Similarly, Hauserman, Zwerback and Plotkin 10 11 (1972) used tokens as reinforcement to show that verbal interactions of adolescents who were state hospital patients could be increased during group therapy. Reinforcement can also be used to increase atten- tion, cooperation, and persistence (Bedner, Zelhart, Greathouse, and Weinberg, 1970). Thus, Specific verbal behaviors, more general behavior patterns both verbal and nonverbal, and the frequency of verbalizations can be modi- fied. Liberman (1970), in a study in an outpatient mental health center, was able, through prompts and verbal rein- forcement, to increase expressions of cohesiveness. Miller and Miller (1970) increased group attendance of welfare families by using concrete reinforcers. Shapiro (1963), using 60 adult women as subjects divided into three—person groups, proved that a type of verbal statement, "disagree- ment," could be increased more by using contingent rather than noncontingent reinforcement. Similarly, Ullman, Krasner, and Collins (1961) used positive reinforcement with a group of male Veterans' Administration hospital inpatients to increase the frequency of "emotional words." Oakes (1962) added another dimension to the rein— forcement research when he conducted a study using flashes of signal light that indicated the subject's statement showed "psychological insight." By using this technique he was able to increase certain verbalizations. 12 In addition to reinforcement, the effect of modeling has been studied by several other behavioral researchers. Kramer (1968) used taped models to increase selected responses in study skill groups. Moore and Sipprelle (1971) obtained results indicating that group subjects observing a model receiving reinforcement for specific verbal statements showed a significant increase in the frequency of similar state- ments. In a like vein, Sarason and Ganzer (1973) tested the effectiveness of using models with groups of delinquent boys. Modeling and imitation resulted in greater short- and long-term changes than did the normal institutional rehabili- tation program. Two other techniques common to behavior modifica- tion are behavioral rehearsal and role playing. Hedquist and Weinhold (1970) compared a behavioral rehearsal proce- dure that included problem specification, role playing, behavioral rehearsal, and in yiyg practice with a social learning procedure that included teaching problem-solving skills. The desired behavior change was assertiveness, and both groups produced a higher level of assertive behavior. Lomont, Gilner, Spector, and Skinner (1969) also demon- strated that assertion therapy taught by means of role playing increases assertive behavior. One study by Heckel, Wiggins, and Salzberg (1962) dealt with negative reinforcement. They used an unpleasant auditory stimulus in a psychotherapy group as a negative l3 reinforcement to decrease the amount of silence in a group. Periods of silence were significantly decreased. Six studies were concerned with the effect of punishment in groups. Studies by Aikin (1965); Bachrach, Candland, and Gibson (1961); Hastorf (1965); and Tyler and Brown (1967) failed to demonstrate that punishment was effective. However, a unique research effort by Levinson, Ingram, and Azcarate (1967) found that the group itself could be used as punishment. Inmates at a boys' training school could earn their way out of the group by not receiv- ing misconduct reports. The group was aversive in the sense that the boys saw the group as useless; therefore the group was the punisher. The Tyler and Brown (1967) study con— trasted "time—out" with punishment and found that "time-out" was shown to decrease misbehavior, whereas verbal reprimands did not. To summarize the Operant behavior modification lit- erature, positive reinforcement, negative reinforcement, shaping, modeling, behavioral rehearsal and role playing, and punishment have been proven to be possible techniques for group treatment. These techniques have been used to influence the frequency and duration of verbal behavior of group members, the type of verbal behavior of group members attending group sessions, group cohesiveness, and group participation. 14 Two social work models of group treatment were offered in the literature. Sheldon Rose (1967) presented a model for group work with children. The second model is the one provided by Lawrence and Sundel (1972), which incorporates a number of behavioral techniques into a spe- cific sequential group procedure. The Rose (1967) and Lawrence and Sundel (1972) studies are unusual in that they tested a comprehensive behavior approach to group treatment and also suggested utilization of group dynamic principles if they further individual goals. The way these authors incorporated punishment was to teach its limits and present its side effects. Group Rational Emotive Therapy Literature Albert Ellis saw Rational Emotive Therapy as being compatible with behavior modification (Ellis, 1973a). The clearest explanation of Rational Emotive Therapy is contained in the popular work by Ellis and Harper, A Guide to Rational Living (1961). Another concise article explaining Rational Emotive Therapy is that by Albert Ellis (1973b) in Psychology Today. Three Rational Emotive Therapy studies are of par- ticular interest, in that they involve the use of R.E.T. in group situations. A study by Meichenbaum, Gilmore, and Fedoravicius (1971) compared R.E.T. to group desensitization and found both to be equally effective for decreashmganxiety. 15 Trexler and Karst (1972) compared the effectiveness of R.E.T. and relaxation techniques for the problem of speech anxiety. Their results indicated that R.E.T. was more effective than relaxation. McClellan and Stieper (1973) conducted a pilot study in group marriage counseling using a combination of three techniques--programmed instruction, R.E.T., and psychodrama--with clients of an outpatient Veterans' Administration clinic. They found that group members lessened anxiety and stress about specific problems and increased their positive communication. There was, how- ever, continued concern about sex, money, and child manage- ment, but the level of anxiety in these three areas had decreased. Limited research has been conducted using Rational Emotive Therapy in groups. The present research is an effort to add to the current body of literature on this subject. CHAPTER IV METHOD Hypotheses The hypotheses for the study are as follows: Hypothesis la: 1b: Hypothesis 2a: 2b: Hypothesis 3a: 3b: Research-—Behavior modification plus Rational Emotive Therapy in outpatient treatment groups significantly increases behavioral change for clients as compared to behavior modification treatment groups alone. Null—~No difference exists in the level of behavioral change between clients in out- patient treatment groups who receive behavior modification plus Rational Emotive Therapy and those who receive behavior modification only. Research--Behavior modification plus Rational Emotive Therapy in outpatient treatment groups significantly increases rational thinking by clients as compared to behavior modification treatment groups alone. Null—-No difference exists in the level of rational thinking between clients in out- patient treatment groups who receive behavior modification plus Rational Emotive Therapy and those who receive behavior modification only. Research-—Clients in behavioral and rational emotive outpatient treatment groups will be able to generalize by applying behavioral problem-solving methods to other situations. Null-~No difference exists on the measures of behavioral problem solving on pretests and posttests for clients in outpatient behavioral and rational emotive treatment groups. 16 17 Operational Definitions Behavior modification refers to the Lawrence and Sundel (1972) model of behavioral group treatment. This model includes: 1. establishment of protreatment norms such as group attendance, 2. enlistment of each group member's participation in aiding other group members in problem specifi- cation and skill practice, 3. selection by the therapist of interventions based on a planned curriculum that outlines the sequences of behavior expected of group members, 4. evaluation each week for the purpose of estab- lishing proximate goals for the next week, and 5. inclusion of group maintenance goals in the planning. 6. Concepts taught to group members are: a. problem specification, b. antecedents, behavior, and consequences, c. selective reinforcement, extinction, punish- ment, time-out, and modeling, and d. baselining, which is the counting and record— ing of the problematic behavior by the client. 7. The format of the sessions is: a. sharing of goals for the session, 18 b. teaching of a particular concept, i.e., problem specification, c. use of examples to clarify concepts, d. use of role plays, actual examples, or beha- vioral rehearsal to make the concepts applic- able to the problems of group members, e. review of behavioral assignments, and f. giving of new behavioral assignments. Worker techniques include mini—lectures, demonstra- tions, illustrations, and feedback throughout group sessions. Handouts are given defining behavioral terms and explaining giving and receiving positive evaluations and criticisms and making requests (Walter, 1973). Walter (1973) defined the constituent techniques for behavioral groups to clarify the means as: 1. Behavior re-enactment: a technique to obtain response display of a target behavior for assessment or modification (via feedback). This technique involves verbal interchanges exemplifying a previous behavioral situation or one similar to a previous situation and may employ either client-client or client-therapist dyads. 2. Behavioral skits: A technique to demonstrate beha— vioral events for illustrating behavioral principles, appropriate v. inapprOpriate behaviors, or appropriate behaviors (such as in modeling). 3. Behavior rehearsal: A technique involving instruc- tions to the client to exhibit in the presence of the group a desired behavior. The purpose of rehearsing a behavior may be to bring that behavior under discrimi- native control, to shape the form of the behavior, to strengthen an infrequent behavior or all of these. 4. Corrective feedback and instructions: A technique combining a discrimination training procedure, using verbal contrast or designations of one or more beha- viors, and behavior prescription. Corrective feedback 19 may occur alone, (the group may be used in providing corrective feedback alone) and instructions may be aimed at in-session responding or extra—session respond- ing or both. 5. Instigation: Prescribing certain behaviors the client is to exhibit in his natural environment (p. 2). An operational definition of Rational Emotive Therapy is the teaching of 11 irrational assumptions and the A, B, C, and D of R.E.T. These concepts were taught by giving clients printed handouts, using mini-lectures, reinforcing rational behavior, and challenging irrational behavior. The handout follows: Irrational Assumptions It is a dire necessity for an adult to be loved or approved of by almost everyone for virtually anything he does. One should be thoroughly competent, adequate, and achieving, in all possible respects. Certain people are bad, wicked, or villainous and that they should be severely blamed and punished for their sins. It is terrible, horrible, and catastrophic when things are not going the way one would like them to go. Human unhappiness is externally caused and that people have little or no ability to control their sorrows or rid themselves of their negative feelings. If something is or may be dangerous or fearsome, one should be terribly occupied with or upset about it. It is easier to avoid facing many of life's diffi- culties and self-responsibilities than to undertake more rewarding forms of self-discipline. The past is all important and that because something once strongly affected one's life, it should defi- nitely do so. People and things should be different from the way they are and that it is catastrOphic if perfect solutions to the grim realities of life are not immediately found. 20 10. Maximum human happiness can be achieved by inertia and inaction or by passively and uncommittedly "enjoying oneself." 11. A person should be rational about almost everything he does or feels. A B C Theory A (Situation) —- B (What I tell myself —- C (How [actcnrfeel) about the situation) (Ellis and Harper, 1961) The ABCD theory is a method by which a person can evaluate the actual situation, what he tells himself about that situation, which is apt to be the controlling variable, and his own action. The D, which was used in the group but not covered in the handout, stands for dispute. The client is taught to dispute irrational beliefs. Outpatient treatment groups designate clients who have come for help with a problem and who are not confined to an inpatient unit. The clients are living on their own. Rational thinking is defined as successfully avoiding irrational assumptions in the process of decision making. Selection of Subjects Subjects selected were regular clients at Ingham Medical Mental Health Center, which serves a client popula- tion over 18 years of age. Ingham Medical Mental Health Center provides service to the southern portion of Lansing and Ingham County. Clients tend to be in a middle or lower income range, and more females than males are seen for treatment. 21 The subjects were selected at random from those who called in for an appointment during a three-week interval in October. Those who agreed to participate were then randomly assigned to one of three treatment groups. All but two clients who showed up for the first interview agreed to participate. Of the 36 contacted by telephone, 16 arrived for the first interview. Subjects excluded from theresearch were active psychotics, substance abusers, and those with marital problems. The rationale for exclusion of active psychotics is that treatment could not easily be confined to a once-a-week group session. Short periods of hospital- ization are often utilized by the case manager, and during an acute episode the client may be seen daily. Thus it would be difficult to separate the impact of group treatment from other therapeutic contacts. Substance abusers were ruled out because other means of treatment are available under the administration of the Tri-County Mental Health Board. Per- sons with marital problems were excluded because it is method— ologically unsound to compare dyads in a group to individuals in a group. No attempt was made to select clients who were more highly motivated than other clients. Subjects were the regular clients who came to the center for help, which resulted in heterogeneously composed groups. Of the 14 subjects who finished the study, the mean age was 36; four were males and ten were females. The edu- cational range was from a high school drOp—out through a 22 person holding a doctorate degree. SeveralVfinxeprofessionals, two were factory workers, several were clerical workers, one was a government employee, one was unemployed, and sev- eral were housewives. Independent Variable The independent variablevuusmethod of group treat- ment. Treatment one consisted of behavior modification according to the Lawrence and Sundel (1972) model. Treat- ment two consisted of behavior modification according to the Lawrence-Sundel model plus the addition of a Rational Emotive Component. Dependent Variables One dependent variable used was a multiple baseline. Each subject was asked to baseline has problematic behavior from the time of the initial interview to the first session (usually one week) and again one week after the last group session. To get some measure of control, the Client was also asked to baseline a second problematic behavior that was not specifically dealt with in the group sessions. A second dependent measure was a client self-report instrument developed and used by Lawrence (1973) in his current research (see Appendix A). This instrument pro— vides data on the problems selected for both baselines. In addition, it asks about relationship changes and also asks 23 the subject to rate the group and different aspects of the group. Another dependent measurevunsa set of three vignettes developed by the Lawrence team with questions designed to elicit whether or not a client can apply the concepts taught in the group (see Appendix A). Two newly developed instruments were used to measure change in rational thinking. The first, the Rational Beha- vior Inventory, was deveIOped by Shorkey at the University of Texas and Whiteman at Michigan State University (1973) (see Appendix A). This instrument provides both an over—all index of rationality and subscores on 12 rationality factors. The present study used the over-all index. The Rational Behavior Inventory was originally tested on university stu- dents who had a mean score of 29.5. Each factor was measured by a Guttman scale with a coefficient of reproducibility of .90 or greater. The total test had a Split-half reliability of .73 using the Spearman-Brown formula. The second instrument was developed by Curran and Collins (1973) (see Appendix A), and was originally used for three groups, which included teachers, students, and helping professionals. This scale is in the process of being further refined. Since both of the measures of rationality are in the developmental stage, results using them are preliminary. The final dependent measure was the therapist rating scale developed by Curran (1974) (see Appendix A). 24 Design Thirty-six subjects were randomly selected and randomly assigned to one of the two treatment conditions or a third group, which was the Rational Emotive Therapy plus behavior modification condition. However, the third group was not protested, with the exception of baselining the two problems. The purpose of the third group was to check the effect of the pretests. All three groups were conducted by an experienced social group worker trained to use both the Lawrence and Sundel (1972) model and the Rational Emotive Therapy plus behavior modification model. The therapist conducting the grOUps had no special interest in one treatment condition over another. Procedure Initial Contact The client contacted the referral secretary by tele- phone or in person. The secretary then determined the nature of the problem and obtained demographic information from the client. The fee was set and the general agency procedure explained to the client. It was at this point that psy- chotics, substance abusers, and those with marital problems were screened out. Of those subjects remaining, a table of random numbers was used to randomly select and then randomly assign subjects to one of the groups. 25 Telephone Contact The name and telephone number of the potential client were given to the therapist, who called to schedule an interview. At this point the clients were usually told they would be receiving group treatment. Initial Interview The therapist spent approximately an hour interview— ing clients. He used the following checklist, adapted from one developed by Harry Lawrence (1973). The therapist was free to change the order of the interview, depending on the circumstances the client presented. Initial Interview Checklist 1. Introductory amenities 11. Overview of interview A. Identify problem B. Nature of groups C. Two assessment activities D. Research component III. Nature of group service A. Time schedule 2nd interview individual--pretests 3rd starts group 6 sessions, 2% hours each 1 or 2 follow—up sessions 26 B. Group sessions 5-7 people training in personal management of problems through group interaction and direction by group leader to learn method of problem solving to apply to other problems focus only on problem agreed to no pressure to discuss history or other personal things C. Group rules attend all sessions no socializing written and behavioral assignments —— ____IV. Verbal statement of client interest A. Questions ____8. Interest in continuing V. Problem inventory and selection ____A. List with examples ____B. Selection VI. Specification A. Description and examples B. Prebase——frequency, magnitude, and duration VII. Baseline plan A. Importance of recording B. Forms C. Demonstrate and practice —— 27 D. Ask C if task is reasonable E. Phone (arrange to phone C and give number where you can be reached) VIIL Research component A. Collecting information about group B. Report for others to use method C. No deception D. Absolute confidentiality, no way to identify people in scientific reports IX. Sign consent forms X. Pretests XI. Overview of next session A. More assessment B. Written assignments will be discussed C. Get C commitment to bring data D. 3rd session will start group XII. Arrange appointments A. Next one B. Group meeting times In the case of those subjects entering the Rational Emotive Therapy plus behavior modification treatment groups, a brief explanation regarding the nature of the groups was added. Second Interview A second interview was conducted to administer the pretests “U3 those receiving them. 28 The Six Group Sessions During the six sessions the worker taught concepts by defining them and giving handouts with the definitions, gave and reviewed behavioral assignments, used behavioral skits and behavioral re-enactment, used behavioral rehearsal, cited and elicited examples, and reinforced goal-oriented activity by group members. In the behavior modification treatment condition the following procedure was used: Session l.--Baselines were reviewed, behavior speci- fication was taught, and new assignments were given. Speci- fication included questions such as who, when, where, and how often. Session 2.--Assignments were reviewed; the concepts, "antecedents,' and "consequences" were taught; and new assignments were given. Session 3.--Assignments were reviewed; positive reinforcement, punishment, and avoidance behavior were taught as concepts; and new assignments were given. Session 4.--Assignments were reviewed, ways to increase desired behavior while decreasing undesired behavior (extinction and differential reinforcement) were taught, and new behavioral assignments were given. Session 5.--Assignments were reviewed, methods of giving and receiving positive evaluations and making requests were taught, and new assignments were given. 29 Session 6.--Assignments were reviewed, giving and receiving criticism was taught, and group members were asked to keep the final baseline for the posttest session. All concepts taught in the six sessions were reviewed. In the behavior modification plus Rational Emotive Therapy sessions, all of the above were included session by session. In addition, during the first session the Irra- tional Assumptions and A, B, C concepts of Rational Emotive Therapy were handed out. This handout was briefly explained and examples were given. In the second session the D (dis— pute) of Rational Emotive Therapy was explained and examples were given. In all sessions rational thinking and challeng- ing of irrational thinking were reinforced. The use of rational thinking was included in the behavioral assignments. All group sessions were recorded and the researcher listened to randomly selected interviews to verify that the treatment schedule was followed and that each treatment condition followed as preplanned. Final Interview The final interview was held to administer the post- tests and collect the final baseline data. CHAPTER.V RESULTS The results are organized into seven major sections. The first section contains the results of individual pre- and postgroup baselines of problematic behaviors. The second section provides a comparison of group results of analyses of variance on baseline behaviors when the baseline scores were converted to an interval scale. Following next are the analyses of variance for measures of rational thinking. The fourth section reports analyses of variance for generaliza- tion of behavioral problem-solving skills. The next part looks at analyses of variance of client reports and esti- mates of improvement by the therapist. The second through the fifth sections report analyses of variance on the depen- dent measures by pair comparisons rather than over-all comparisons. The sixth section presents correlations between selected dependent variables. The final section provides a report of whether or not the therapist followed the prescribed treatment regimen for each group. Baseline Behaviors Each client selected one and if possible two proble- matic behaviors to measure. The first problem measured was 30 31 subject to modification in the group and the second one was not subject to modification by group procedures. The client wrote down the baseline information. In some cases the client forgot or wrote a narrative so that it was necessary for the therapist to make an estimate based on a conversa- tion with the client. The individual baselines for group one, which was the group receiving behavior modification, Rational Emotive Therapy pretests, and posttests, follow: Client one was a 33-year-old female in the process of obtaining a divorce. She was in a middle-income range. Figure 1 shows the behavior that was modified in the group. The goal was to decrease the number of periods of anxiety each week. Figure 2 shows the baselines of the second behavior, which was not dealt with in group sessions. The goal was to decrease the frequency of angry outbursts each week. For this client both negative behaviors decreased. W l o u m 10 8 15 '8'?» M. "4 a) 8 ‘ H G) 12 --1 L43 7‘ 0‘9 O C 3 an 6 m 9 m u “a e .. 8. >‘ 4 0 6 238 “ 1’3 .9... 2 8 m 3 63¢ E H :3: s 5 Z '5 0 r1 2 n 0 II Before After Before After group group group group treatment treatment treatment treatment Figure l.--Number of periods of Figure 2.-—-Number of angry anxiety per week. outbursts per week. 32 Client two was a 32-year-old male who works for the federal government. Figure 3 reports the results of the problematic behavior worked on in the group, which was to decrease the number of times he presented himself as a loser. Figure 4 reports a second goal, which was not worked on in the group. He wanted to increase the number of times he gave behavioral criticism rather than "name-calling"cmiticism. Mel 7" Mia mczu 15 we: 5 QAJO (LA F— m‘é’IS mU whhd 12 m 4 5% ES pr: 9 +ba 3 cm > ‘Hdlm uamwj om4 6 orzm 2 v-HH (DU) acne urns m m 3 0 1 .gacm .Q#:E :85 58?: 2:35: 0 F1, 2:30 0 F1 Before After Before After group group group group treatment treatment treatment treatment Figure 3.--Number of times per Figure 4.--Number of times week client presented himself per week behavior criticism as a loser. used. The client achieved change in the direction desired in both cases. Client three was a 28-year-old woman who had recently been divorced. She was also employed and was in a middle- income range. Figure 5 shows the behavior worked on in the group, which was to decrease the periods of depression per 33 week. Figure 6 shows the second problematic behavior, lack of assertiveness. Figure 6 reports an increase in assertive behavior. The client achieved change in the desired direc— tion in both instances. w 12 12 °f 3 mo) 10 __ Q, 10 "03 _— .943... 8 8% 8 um ES ,8.“ 3.. c 6 c 6 “5.9. “5.22 m 4 F+H 4 Lem u OLA QIw m u DH 2 am» 2 En: ECHO 5a) 9 man 21: 0 IN 2 3n: 0 F1 Before After Before After group group group group treatment treatment treatment treatment Figure 5.--Number of periods Figure 6.--Number of times of depression per week. per week client was asser- tive. Client four was a 25-year-old woman who was trying to increase her assertive behavior. She was not able to select a second behavior. Figure 7 shows her baselines. Change occurred in the desired direction. 34 Number of times per week client was assertive OJ 2 l 0 I Before After group group treatment treatment Figure 7.--Number of times per week client was assertive. Client five was a 27-year-old woman employed as a clerical worker, who wished to increase her assertive responses to her boyfriend. Her second problematic situa- tion was that she wanted to decrease the frequency of depen- dent actions. Figure 8 shows the behavior worked on in the group and Figure 9 shows the control behavior. Change occurred for both behaviors in the desired direction. Client six was a 21-year-old female student. Her problem chosen for group treatment was to increase uninter- rupted study periods. Her usual pattern was to daydream while she was supposed to be studying. She did not provide a second problematic behavior. Figure 10 gives the results for this client. Change was in the desired direction. 35 I H m u 14 as: s more 10 ex 12 — woes '— $3; (Um-H 8 9,3 10 Silk m -a 44 ps4 we)» 6 m 8 :30 qu. see 0. ~40 4 :4c 6 HEJU an m Qw4 beam 2 E+l 4 Ed)> SI) Dara qu 253*) 0 r3_ 2 , - Before ‘After Before After group group group group treatment treatment treatment treatment Figure 8.--Number of times Figure 9.-—Number of depen- per week Client was asser— dent actions per week. tive to boyfriend. x m es lo 15 ALCH SIMS. —- CF! 12 view :Lce SIC) 9 ea wIcu c>om 6 Q HTS (DCD>~I 3 Line 50.: fl snap zxun 0 ,, Before After group group treatment treatment Figure 10.--Number of uninter- rupted one-half hour study periods per week. 36 To summarize, clients in group one all achieved changes in both the target behavior and a second behavior, if one was selected. Results now are given for group two, which received behavior modification only along with the pretest and post- test. The seventh client cited was a 57—year—old woman who was separated from her husband. She would have liked a reconciliation, but was extremely passive. The problem she identified was to increase her assertive behavior. She did not specify a second problem. Figure 11 gives the baseline data. She increased her assertive behaviors from one to three. Number of assertive behaViors per week : II Before After group group treatment treatment Figure 11.——Number of asser- tive behaviors per week. 37 The eighth client was a 54-year-old, overweight housewife who wished to change her eating behavior. Her husband was a professional man with higher than average earnings. A second set of baselines could not be obtained. Figure 12 shows she decreased the snacks she usually had on shopping trips. 10 $2 0 8 F! m If, 6 £64 Iafi 4 um 83 EM 2 28. 0 II Before After group group treatment treatment Figure 12.——Number of snacks per week. The ninth client was a 23-year-old unemployed male who lived with his elderly parents. He had relatively few contacts with women. He had a second goal related to sexu- ality, but it was a long-range goal that was inappropriate for a second set of baselines. Figure 13 cites approach responses to women. Minimal change occurred for this client. 38 a c 5 o e 8% p 3 4 8 m.8 3 w m o. 5 I— M10 Si? N mIm 4 .§:3 8 e3 an: ._ r683 __ roe. 6 ‘8‘” m m —a m 2 0t; 4 ea W .orol 35 2 E-H E m 234-) 55-4 2 m 0 2:6 0 Before After Before After group group group group treatment treatment treatment treatment Figure l4.--Number of anxiety Figure 15.--Number of dish- attacks per week. washing arguments per week. at the time of group termination. The client was a 33-year- old employed male. Figure 16 shows the decrease in depressive episodes. : .9, 10 314 e e 8 u e , 2:3 ‘I e u 6 e “50* 4 t8 an 2 Q-H ea: Fl 2 m 0 . . Before After group group treatment treatment Figure l6.—-Number of depression periods per week. 40 As in group one, clients in group two did report change in the desired direction in all behaviors reported. The final group received both behavior modification and Rational Emotive Therapy. However, only posttests were given with the exception of baseline material. Only three subjects completed this group. This failure to complete the treatment sequence will be covered in the discussion section. The twelfth client of this research was a 57-year-old unemployed female. Figures 17 and 18 show the results for this client. Figure 17 indicates a decrease in staying in bed all day. Figure 18 depicts increased verbal assertive- ness. $13 5 5 QC) :1: S meew 4 4 ems UL) 3 —- 3 .— CI—I “5.273 A 2 2 sue QALQ 1 1 sec fl 2 3H 04_ Q 0 ' . Before After Before After group group group group treatment treatment treatment treatment Figure l7.--Number of times per Figure l8.-—Number of times week client stayed in bed all per week client was ver- day. bally assertive. 41 The next client was a 26—year-old, recently divorced female who was coping with the effects of the divorce. Figures 19 and 20 show the results for this woman. Her avoidance of people was charted in Figure 19 and her initia- tion of conversations in Figure 20. Both show a change in the desired direction. 12 10 r ' 5 8 4 6 3 _ 4 2 1 2 l l 0 0 7| Before After Before After group group group group treatment treatment treatment treatment Figure l9.--Number of times Figure 20.--Number of con- per week client avoided versations initiated per people. week. The last client reported was a 59-year-old housewife whose major problem was anxiety attacks. She was the wife of a professor. The numbers of pre- and postgroup anxiety attacks are shown in Figure 21. Figure 22 shows the number of times the client judged she was behaving in a rigid man- ner. No change occurred in rigidity. A summary of group three indicates that change occurred in all but one instance. 42 18 >« 15 F1 (nu UM '00 10 .38 12 3'; :3 ”I are 8 (ULI 9 am) FT "— m D 6 m Q w o owLx u) 6 u-Ia) u.x were 4 mt) Ow13 etc 3 £34 g.p E H 2 g.u SWIG z m 0 zc>m 0 Before After Before After group group group group treatment treatment treatment treatment Figure 21.--Number of anxiety Figure 22.--Number of per- attacks per week. iods of rigid behavior per week. Analyses of Variance of Baseline Behaviors To provide comparisons among subjects and between groups, individual scores were converted to an interval scale similar to the Goal Attainment Scaling (see Appendix B) used in several services at Ingham Medical Mental Health Center. Each client reported a criterion level for desired change. The scale is indicated as follows: 1 = No change 2 = Less than criterion level 3 = Criterion level 4 = Improvement above the criterion level 5 = Elimination or the highest level of achievement Each client was given a rating based on his stated criterion level. Table 1 summarizes the analyses of variance 43 between the three treatment conditions. The significance level was set at .05. Table l.--Ana1yses of variance of baseline behaviors. «m__. -I- _.__~..-....._...__ - . . - .....i . _ -___ _--__. __._'-__ _ _. ___~__.__._—- _-__ -.._..._ M..... - . -- -.. , ____.-__ -7. - __.__._. _ _ . —._—.__ _... —-.__ _.—____. Degrees of Com arison Means p Freedom Behavior modification plus (a) (b) R.E.T.(a) with behavior modi- fication on1y(b). Both 4.50 2.80 1,9 8.55* received pre- and posttests. Behavior modification plus (a) (c) R.E.T.(a) with behavior modi- fication plus R.E.T. with no 4.50 3.33 1,7 3.09 pretest(c). Behavior modification only (b) (c) with pretest(b) and beha- vior modification plus 2.80 3.30 1,6 .42 R.E.T. without pretest(c). *Significant at .05 level. The difference that was significant was between the behavior modification plus Rational Emotive Therapy condi- tion and the behavior modification only condition. The other two comparisons were not significant at the .05 level. The research hypothesis was supported by the first comparison and the null hypothesis by the third comparison. Analyses of Variance of Measures of Rational Thinking It was predicted that clients in the rational therapy treatment condition would receive a more rational score than 44 those who were not in that treatment condition. In order to determine this, F ratios were computed for the pre- and posttests for the behavior modification plus Rational Emotive Therapy condition and also for the behavior modification only condition. A posttest comparison was computed to mea- sure differences between groups. Table 2 shows the F score for pretest-posttest differences on the Rational Behavior Inventory for the behavior modification plus Rational Emotive Therapy condition and also for the behavior modification only condition. A higher mean indicates a more rational score. Table 2.--Analyses of variance for the Rational Behavior Inventory. Degrees of Comparison Means Freedom F Prebehavior modification _(a) (b) R.E.T.(a) with postbehavior . . . . . .,- . 7 modification plus R.E.T.(b) 20 67 23 50 I 10 2 9 Prebehavior modification(c) (c) _(dl_ With postbehaVior modifi- 25.20 25.40 1,8 .004 cation(d) The F scores were not significant. Table 3 provides the posttest score comparisons between all three groups. The F scores were not significant. The posttest scores are less meaningful because of the pretest differences between groups. An F for gain scores was computed for 45 groups (a) and (b). An F of .02 was obtained and was not significant at an .05 level. Table 3.--Posttest analyses of variance between groups on the Rational Behavior Inventory. _---_.-.. .--.. .- .__-._-- .._<-_ .._—_- .-L. -. . _.__. ._- - ‘_ . ... . ..__ -- . v..._ _. -.-..--.7.— Degrees of Freedom F Comparison Means Behavior modification plus R.E.T.(a) with behavior mod- (a) (b) ification on1y(b). Both 23.50 25.40 1,9 4.66 received pre- and posttests. Behavior modification plus R.E.T.(a) with behavior mod- (a) (C) ification plus R.E.T. with 23.50 28.30 1,7 .621 no pretest(c). Behavior modification only (b) (c) with pretest(b) and behavior modification plus R.E.T. 25.40 28.30 1,6 2.33 without pretest(c). The second test used to measure rational thinking was that develOped by Collins and Curran (1973). Tables 4 and 5 are the analyses of variance for the Collins and Curran scale. Table 4 compares pre— and posttests for the two treatment conditions. A lower score indicates more rational thinking. Neither F was significant. Posttests were compared in Table 5. No significance was found. A gain score was computed with an F of .00, which was not significant. 46 Table 4.—-Comparison on the Collins—Curran scale--analyses of variance. ___—___.._. .- --_*____~ -._.7n..._._._. _- , ._V‘ -.__.. Degrees of Com arison - Means p Freedom Prebehavior modification (a) (b) plus R.E.T.(a) with post- behavior modification 49.50 45.50 1,10 1.06 plus R.E.T.(b) Prebehavior modification(c) (c) (d) with postbehavior modi- fication(d). 44.40 43.80 1,8 .02 Table 5.--Posttest comparisons between groups on the Collins- Curran scale. Degrees of Comparison Means Freedom F Behavior modification plus R.E.T.(a) with beha- (a) (b) vior modification only(b). Both received pre- and 45.50 43.80 1,9 .09 posttests. Behavior modification (a) (c) plus R.E.T.(a) with beha- vior modification plus 45.50 45.33 1,7 .01 R.E.T. with no pretest(c). Behavior modification only (b) (c) with pretest(b) and beha- vior modification plus 43.80 45.33 1,6 .10 R.E.T. without pretest(c). 47 To summarize, neither the Rational Behavior Inven- tory nor the Collins-Curran scales of rational thinking showed significance on any of the measures reported. Thus the second hypothesis was not supported. Those subjects receiving Rational Emotive Therapy did not respond more rationally on the rational thinking scales. Analyses of Variance for Generalization of Behavioral Problem Solving One of the research questions raisedefiswhether or not clients could extend their knowledge of behavior modi- fication problem—solving methods to new situations. This was tested in a series of three vignettes deveIOped by Lawrence (1973) dealing with different types of problem situations. Each vignette has a range of possible scores from 0 to 7. Vignette 1 deals with a marital problem. Vignettes 2 and 3 depict child-related problems. Table 6 provides a comparison of pre— and posttests for all three vignettes for the behavior modification plus Rational Emotive Therapy and treatment and the behavior modification only treatment. None of the F ratios for the vignettes produced significant scores at .05, but the pre-post on Vignette 3 of the R.E.T. group was one of the higher P scores of this research. 48 Table 6.--Analyses of variance of pre— and posttest comparisons on generalization vignettes. . _._. -_.____—_-_—_.._—._..__—.—~._--_- __.___- - , _ ., Degrees of Comparison Means Freedom F Prebehavior modificationrfliu; Pre Post R.E.T. with postbehaviorrmxl— -——- ———— ification plus R.E.T. on 3.17 4.67 1,10 .52 Vignette 1 Prebehavior modification Pre Post with postbehavior modifi- cation on Vignette l 2'60 4°60 1'8 32 Prebehavior modification Pre Post plus R.E.T. with post beha- ———— ———— vior modification plus 4.50 4.83 1,10 .53 R.E.T. on Vignette 2 Prebehavior modification Pre Post With postbehaVior modifi- 3.20 3.60 1,8 .084 cation on Vignette 2 Prebehavior modification Pre Post plus R.E.T. with postbeha- ———— ———— vior modification plus 1.33 2.50 1,10 .62 R.E.T. on Vignette 3 Prebehavior modification Pre Post With postbehaVior modi- 4.20 5.00 1,8 .31 fication on Vignette 3 To test the generalization hypothesis between groups, analyses of variance were employed. Table 7 summarizes the generalization findings for Vignette 1. used. No significance was found. A level of .05 was Table 8 charts the same information for Vignette 2. None of the measures proved significant at the .05 level. 49 Table 7.-—Ana1yses of variance posttest group comparisons on generalization Vignette l. Degrees of Com arison Means' p Freedom Behavior modificationgflxus (a) (b) R.E.T.(a) with behavior mod— ification only(b). Both 4.67 4.60 1,9 .00 received pre— and posttests. Behavior modificationpdus R.E.T.(a) with behavior mod- (a) (C) ification plus R.E.T. with 4.67 4.33 1,7 .03 no pretest(c). Behavior modificationcnflqr (b) (C) with pretest(b) and behavior modification plus R.E.T. 4.60 4.33 1,6 1.84 without pretests(c). Table 8.--Analyses of variance posttest group comparisons on generalization Vignette 2. _ . ‘ .-_‘___.--__._.—-—‘_-.—__ - . a...» .--_.-_ Degrees of Freedom F Comparison Means Behavior modification plus (a) (b) R.E.T.(a) with behavior modification only(b). Both 4.83 3.60 1,9 .62 received pre-enuiposttests. Behavior modification plus (a) (C) R.E.T.(a) with behavior modification plus R.E.T. 4.83 6.00 1,7 .51 with no pretest(c). Behavior modification only with pretest(b) and beha- vior modification plus 3.60 6.00 1,6 1.84 R.E.T. withoutpmetests(c). 50 Table 9 summarizes the generalization findings for Vignette 3. Table 9.--Analyses of variance of posttest group comparisons on generalization Vignette 3. —_..___._—_—__. ._-.____ »- . _.__.__. —. .-. _. ._.___-.__—._..__.__._..__ _..-‘_..._——.__.__-_-.__‘_ _. ._._. __ _ __._.-_._._a __.._._.. _ .____._.... —_.__...___._..-..w._._.._.._ .EM-.~.-.~..WWW—pflwmfla__.-.m._—..._i W-.HW-.L_1-... ....L. ._ ..._ . is..._A-___.__—..._.... -..._....._......_.._.... -kw. _._...___. Degrees of Com arison Means p Freedom Behavior modification plus (a) (b) R.E.T.(a) with behavior modification only(b). Both 2.50 5.00 1,9 8.77* received pre- and posttests . Behavior modification plus R.E.T.(a) with behavior (a) (C) modification plus R.E.T. 2.50 2.67 1,7 .05 with no pretest(c). Behavior modification only with pretests(b) and beha- vior modification plus 5.00 2.67 1,6 3.28 R.E.T. without pretests (c) . *Significant at .05 level. Results indicate that the F of 8.77, which compares the behavior modification plus R.E.T. with behavior modi- fication only, was significant at the .05 level. Gain scores were computed for all three vignettes; none was significant at .05. A summary of the vignettes as a test of generaliza- tion indicates only one instance of significance, which was on Vignette 3. The three vignettes were combined into an index comparing pre- and posttest scores. A t test was 51 computed for both treatment conditions. In the combined treatment the difference was significant at .01 and in the behavior modification group the difference was significant at .025. Analyses of Variance of Client Reports and Therapist's Estimate of Improvement Clients had a choice of ratings from 0—7 on a con- tinuum from very much worse to very much better. Mean scores are reported in Table 10. Table 10.--Mean scores on client problem checklist. _.__..._ I __._.___ - _. ___.___.——....I__ Treatment Mean Group 1-—Both treatment conditions, pretests and posttests 5.67 Group 2-—Behavior modification only, pretests and posttests 5.40 Group 3--Both treatment conditions, no pretests 6.00 The mean scores indicate a high level of change in the Clients' estimate in all three groups Table 11 shows F scores for all three groups. None of the F ratios was significant. A second part of the Lawrence (1973) evaluation was a listing of the number of relationships that could improve during the duration of group treatment. It was possible to 52 Table ll.--Ana1yses of variance of Clients' problem checklists. Degrees of Comparison Means Freedom F Bouitreatments(a) and beha- (a) (b) vior modificathmaonly(b). 5.67 5.40 1,9 .87 Behavior modifidation (b) ( ) only(b) with both treat- C ments (no pretests)(c). 5.40 6.00 1,6 2.57 Both treatments with (a) (c) both treatments but no pretest. 5.67 6.00 1,7 .89 indicate up to 12 relationships had improved. For example, a client indicated he was getting along better with his neighbors, co-workers, and wife. The example would give a score of three. Analyses of variance did not indicate any significant differences. Results are shown in Appendix B. It is interesting to note that the mean for all three groups is 5.1. This tends to suggest some generalization of methods learned in the groups to multiple relationships. A third part of the Lawrence (1973) evaluation asked clients whether or not they needed help on their selected problem or another problem at the conclusion of the group. No significance was found; results are reported in Appen- dix B. The last section of the Lawrence (1973) instrument asked a series of questions about the source of help. No significance was found. Appendix B includes these data. 53 In general, cleints reported favorably on the self- report measures. However, no distinctions could be made between treatment conditions. The final part of this section concerns an evaluation by the therapist on client change for all three groups. Results are cited in Table 12. Table 12.—-Curran therapist rating scale analyses of variance. M..____—,~__ _—.____._ __.— __ _. ._ “_____._._ ...——b———— _. gr ___._.__ Degrees of Freedom Comparison Means F Behavior modification plus (a) (b) R.E.T.(a) with behavior modification only(b). Both 312.00 226.40 1,9 12.43* received pre- and posttests. Behavior modification plus (a) (c) R.E.T.(a) with behavior modification plus R.E.T. 312.00 230.33 1,7 3.57 with no pretests(c). Behavior modification only (b) (C) with pretests(b) and beha- vior modification plus 226.40 230.33 1,6 .00 R.E.T. without pretests(c). *Significant at .05 level. A significant difference at .05 was found between the two treatment conditions. To an extent this supports all three hypotheses. However, the group that was not pre- tested and that received both treatments did not show a dif- ference between itself and the behavior modification only group. That finding does not support the hypotheses. 54 Correlations of Dependent Variables Table 13 presents correlations on relevant dependent variables. Table 13.-~Correlations. Instruments Correlation Collins and Curran and Rational Behavior Inventory pretest -.63 Collins and Curran and Rational Behavior Inventory posttest -.46 Vignette l and Vignette 2 (posttest) .55 Vignette l and Vignette 3 (posttest) .03 Vignette 2 and Vignette 3 (posttest) -.39 It should be noted that the Collins and Curran and the Rational Behavior Inventory correlate in a negative rela- tionship. This is appropriate in that one was scored posi- tively and one negatively. Although the correlations are not extremely strong, there is some evidence of a relation— ship. The other relationship above .5 is that between Vignette l and Vignette 2. Validation of Therapist Interventions To assess whether or not the group worker followed the prescribed treatment regimens, a neutral evaluator was 55 employed. The person selected was a graduate student in com- munications who was trained to distinguish between behavior modification and Rational Emotive Therapy. Each session of a group was audio—taped. From these audio-tapes 33 randomly selected two-minute intervals were rated. The evaluator did not know which treatment condition she was rating. She rated a segment either as R.E.T., Behavior Modification, or neutral. In no case was the therapist rated as doing an inapprOpriate treatment. He was following the specified regimen in 15 instances. In 18 instances clients were talk- ing and it was not possible to determine the treatment from the random time. Summary of Results Although behavioral change occurred in all three groups, there was a significant difference between the beha— vior modification plus R.E.T. and the behavior modification only groups. The combined treatment produced a signifi— cantly higher level of change, as measured on an interval scale. Thus the hypothesis that behavioral change is increased by adding a cognitive component was supported. The Curran (1974) therapist rating scale confirmed this finding. Two measures were used to determine the effects of group treatment on rational thinking——the Rational Behavior Inventory (Whiteman and Shorkey, 1973) and the Collins- Curran Inventory (1973). No significant results were found; 56 however, the Rational Behavior Inventory showed an increase in rational thinking but not at a significant level. Three vignettes with questionnaires were used as tests of generalization. Significant change was found between the behavior modification plus R.E.T. treatment and behavior modification only condition on Vignette 3. The behavior modification only group scored higher. The other two vignettes resulted in no significant differences between treatment conditions. A questionnaire (Lawrence, 1973) was administered to clients following treatment. Included in this measure were a client report on improvement, a count of the number of relationships that had improved, a count of whether the client needed help with his problem or another problem, and a report on the source of help. These measures indicated changes occurred in all three groups but there was no differ- ence between groups. The final measure was the use of an independent observer to evaluate whether or not the therapist was admin- istering the prescribed treatment. In all instances he was adhering to the treatment regimen. CHAPTER VI DISCUSSION The discussion is divided into four major sections. The first relates this particular study to research problems of clinical settings. The second discusses specific results and the implications for each of the three hypotheses. _Section three cites some effects that participation in the research project had on the clients. The final section con- tains suggestions for future research, both clinical and laboratory. Research in Clinical Settings Gordon Paul (1969) listed the four domains of clini- cal research as clients, therapists, time, and criteria. By criteria he meant treatment effectiveness. It is in the client domain that particular problems arise for group research. Sample size is the first client-related problem considered. As noted earlier, group outpatient studies are less frequent than either laboratory or inpatient studies. Of the operant studies reviewed in Chapter III, only seven dealt with outpatient groups. Of the seven, four compared two groups, two examined only one group, and the Lawrence and Sundel (1972) study examined differences in three groups. 57 58 It would be ideal statistically to be able to have enough groups in each treatment condition so that group means rather than individual scores could be used for a part of the analysis. It would be preferable to have six or seven groups in each treatment condition. However, obtaining such a large number of clients is not possible nor feasible without access to a research grant. The present study attempted to provide the largest sample size possible within the bounds of the agency in which the research took place. A second client domain problem previously noted was that clients do not come to an agency as research subjects but rather for help with a concrete problem. This writer participated as a group worker in the current Lawrence research (Lawrence, 1973). An incident occurred that illus- trates the type of frustration a clinical researcher can encounter. A client was trying to increase his verbal com- munications with women in the hope that he could establish a meaningful relationship with a woman in the near future. Eight sessions were slated for the group, but the client eloped and left the state between the fifth and sixth ses- sions. He sent this writer a letter stating he was delighted with the group and his goal achievement. However, he was not available for posttests and thus had limited value as a research subject. In this study the domain of the therapist presented no problems, in that the same therapist had the expertise to conduct all three groups. 59 Time presented a significant problem. Two subjects were lost because of a combination of extremely inclement weather and time. The third group, which received both behavior modification and Rational Emotive Therapy but no pretest, lost the two members, leaving only three in the group. Heavy snowstorms caused the Mental Health Board to close the center on the dates scheduled for the last two sessions of this particular group. The group sessions were rescheduled, but by the time the group met again the two clients were not available. Because of a difficult preg- nancy one woman was too ill to return for treatment. During the same period, the second client had been called to another state to be with a seriously ill family member. The loss of these subjects plus the time lapse caused by the reschedul- ing make the results for this third group much less meaning- ful. Because of the small n remaining, the effect of the pretest was not determined. One problem occurred in measuring treatment effec- tiveness. As noted earlier, in a few instances clients did not baseline. The problem arose when the client did not return the requested frequency counts but rather brought back a narrative which, in the client's eyes, indicated growth or self-discovery. The baseline figure was then obtained by the therapist after a discussion with the client regarding the frequency of the problematic behavior. He made as accurate an estimate as possible and asked the client 60 to verify his figure. Because the self-report method of baselining is viewed with skepticism by some researchers, behaviorists are making increased use of trained observers for baselining. It is probable that a trained observer observing the client in his actual life situation could have provided more precise data, but such observation presents several complications for clinical researchers. First, no set of observers is readily available to be with clients outside of treatment groups. Second, such an observer might be neither desired nor tolerated by at least a portion of the clients. (This could be particularly true if they had watched the 1973 television special series that followed the life of an American family. There have been some indi- cations that the ubiquitous presence of the observer might have affected the Loud family, even to the extent of facil- itating the dissolution of the family unit.) In addition, if someone else were to baseline, the therapeutic value of baselining would be lost for the client. Patterson and Gullion (1967) indicated that sometimes the process of client baselining brings behavior change in the desired direction. The clients in this research found the baselining and other required recording both revealing and meaningful. A second problem with baselining was controlled in this study. From a research point of view it would be ideal to have a weekly frequency count of the occurrence of the problematic behavior. However, because behavioral assignments 61 were also used, it was felt that clients might become over- whelmed with paper work. Thus, record keeping was kept within reasonable limits for the clients and baselines were only obtained prior to the group and subsequent to the group. Discussion of and Implications of the Results The individual baselines (Figures 1 through 22) showed impressive behavioral changes for a majority of the Clients. Changed problematic behaviors included: reducing anxiety attacks and depressive episodes, reducing impulsive or angry outbursts, decreasing presentation of self as a loser, increasing appropriate criticisms, increasing assert- ive behavior, decreasing dependency, increasing studying behavior, decreasing snacking behavior, decreasing argu- ments, increasing out-of-bed time, decreasing avoidance of people, and increasing verbal contacts. These changes occurred in all three treatment conditions. The findings indicate that the two treatment modalities are apprOpriate for a wide range of problematic behaviors. The client self- report scores provide verification for this finding. It is interesting to note that change occurred almost as fre- quently on the baselines of the second problem as on the baselines of the target problem. Therefore, the use of the second problem as a control was not effective. It seems likely that clients generalized the problem-solving methods presented in group sessions to the second problem. Wolf 62 and Risley (1971) pointed out the problem of a second base- line: One possibility is that there will be induction from one baseline to the next. That is, the change that a treatment condition seems to produce in the treated behavior may also appear in the second baseline that is intended to act as a control. The fact that change occurs across both behaviors diminishes the usefulness of the second baseline as a control (pp. 316-317). In the current study it seems quite probable that there was generalization of learning between behaviors. In some cases the second baseline selected by the client was actually worked on as a target problem for someone else in the group. For example, client four worked to increase assertiveness while assertiveness was the control behavior for client three. Under these conditions it would be expected that change would occur in the second problem baselined by client three. A second reason for expecting change in the second problem is that in some instances both problems a client cited were interrelated in some way. For example, one client worked on getting out of bed and her second prob- lem was increasing verbal assertiveness. Getting out of bed increased the probability that she would be more likely to be in contact with people, thus increasing the Opportuni- ties for verbal assertiveness. Keller (1963) noted that generalization is more apt to occur between two stimuli when they are physically similar to each other if all other factors are equal. To avoid generalization effects, it would have been preferable to elicit a second problem quite different 63 in as many respects as possible from the first problem. However, as previously noted, clients come with actual prob- lems, not problems specifically tailored for research. The analysis of variance using an interval scale for the baselined behaviors did show a_significant differ- ence between the behavior modification only and behavior modification plus Rational Emotive Therapy treatment condi- tions. The group receiving the combined treatment produced significantly greater changes than those receiving behavior modification alone. The major hypothesis of the study is supported by this finding. The finding is further substan- tiated by the significant differences found between the groups on the Curran therapist rating scale. The fact that the third group, which also received the combined treatment, did not fare as well is not seen as particularly challenging to the findings because the third group lost two-fifths of its membership. It appears from this research that adding a cognitive component to behavior modification does increase its potency. An implication of this finding is that clients for whom there was an urgency for rapid behavioral change could be included in a behavior modification and Rational Emotive Therapy group without any loss of efficiency compared to behavior modification alone. Clients who need to change rapidly in order to protect themselves or someone else rep- resent a significant portion of outpatient clinic popula- tions. Examples include child-abusers, people with behavior 64 that is jeepardizing their continued employment, and those who are seriously violating the law. The second hypothesis in this study predicted that clients in the combined therapy treatment condition would receive a more rational score than those who were not in that treatment condition. Although the results did not affirm this hypothesis at an .05 level of significance, the results on the Rational Behavior Inventory gave an F score of 4.66 between the two treatment conditions. The high F in postscores is accounted for by pretest differences, as indicated in Table 2. The Rational Behavior Inventory used a cut-off point on each item, which determined whether or not the score was judged to be rational. It was noted in this research that many subjects moved one step on the scale from the pretest to the posttest. Such movement did not give them a rational score but it did indicate movement toward rationality. The third hypothesis tested for generalization of behavioral problem-solving methods to new situations, as represented in the vignettes. The third vignette showed greater pre-posttest differences in the combined treatment condition than did the other vignettes. This vignette also showed significant difference at .05 between the behavior modification plus R.E.T. and the behavior modification only group. The behavior modification only group scored higher. One could speculate that the content of this vignette was 65 broad enough for more people to relate to it since it dealt with both marital and child-rearing problems. The scores on this vignette did not correlate with either of the other two. It is probable that the third vignette was a large factor in the significance of the combined index. The cor- relations were .03 and —.39. Vignettes l and 2 dealt with a single issue, and correlated at .55. One could extend the principles of behavioral generalization to state a client can generalize to a new problem and situation to the extent that it is similar to his problem and situation. Therefore, for a single man to generalize problem-solving skills to a child management discussion might be too far removed from his situation. Thus, perhaps the ideal general- ization test would be a vignette designed by the researcher based on data gathered from the actual subjects, including age, sex, marital status, and presenting problem. The vignette could then present a problem within the realm of possibility for clients. Another measure of generalization was a total of the number of relationships on the checklist that had improved during group sessions. Although there were no significant differences between groups, the over-all mean was 5.1 relationships. This is logical in that some of the Clients' goals were across peOple; i.e. to be assertive did not always state a particular person as a recipient of the behavior. Thus for some clients generalization was built in to the treatment. Others improved their relationships 66 with people without being specifically trained to do so. One could argue that there was a modeling effect between subjects. It was pointed out in the results that a trained observer validated the therapist interventions. The thera- pist was employing the apprOpriate treatment in the approp- riate treatment conditions. This is one of the most exciting findings of the research. Because behavior modification and Rational Emotive Therapy can be Operationalized and the Operations can be followed relatively precisely, the research is replicable. Although it is true that each client is unique and each problem specific to the Client, the proce- dures of therapy can be standardized. Such a finding indi- cates that group work can be based on theory and knowledge rather than on the individual style of a particular worker. One of the values of both treatment methods utilized in this research was that they were taught to clients rather than simply practiced on the clients. The importance of this approach is that in a time of future crisis the methods of behavior modification and Rational Emotive Therapy are known and available to the client. Some Additional Effects of Client Participation in the Research From the beginning clients were interested in par- ticipating in the research effort. Comments were made such as "I'm glad they are doing this," and "It is good that 67 people want to know whether or not we are helped." It became apparent that participation in the research had special meaning to the clients involved. One speculation arising from this fact is that the process of asking clients to participate in a rather intensive evaluation restores some of the dignity they may have felt they lost when they became clients. The research process seems to give a direct message to clients that their opinions, feelings, behavior, and change levels are important. The person conducting the treatment groups made the observation that the clients seemed to give more validity to the treatment modalities because of the research effort than they might have otherwise. For example, there was less than the usual amount of grumbling about receiving group treatment rather than individual treatment. Finally, the therapist received a number of positive evaluative comments about the research and the groups. Sev- eral group members indicated they wanted to read the research findings. This is particularly interesting, in that sub- jects were not aware of the comparisons being made. Suggestions for Future Research The first recommendation is that the comparison between behavior modification outpatient group treatment and behavior modification plus Rational Emotive Therapy be replicated by other researchers in mental health centers and 68 family agencies. Such replication is one way to counter the problem of small n's in clinical settings. Replications would have the added value of providing data for reliabil- ity checks on the dependent measures. Since so little research has been done in behavioral and cognitive social group work, most of the dependent measures are newly devel— oped. n A second recommendation for clinical research grew out of the observations of Clients' reactions to the research process. Clients who are participants in a research project could be measured on self-esteem scales among the posttests. Their scores could be compared with clients who did not participate in research but were given a self-esteem scale at the termination of the treatment. A large n could be collected over a period of a year at Ingham Mental Health Center, for example, since at least five or six research projects are conducted each year with the client population at that center. It could be argued that generalization of the problem- solving methods could be improved with specific generaliza- tion training. It would be interesting to compare behavior modification groups with generalization training and beha- vior modification groups without generalization training. Generalization training could consist of elicitation from clients of other problems to which they could apply behavior modification principles. Clients in the experimental group 69 could practice problem specification, identification of antecedents and consequences, reinforcement, and extinction. Generalization research could be conducted in both the lab- oratory and the clinic. Techniques for maximizing general- ization could be refined in a laboratory, then taken to an outpatient clinic population for testing. Generalization is an important issue since a large number of clients in outpatient settings have been clients previously (Van Westen, 1974). If they were able to generalize their learning to new situations, it would not only be of immeasurable benefit to the client, but would also provide for a more efficient use of mental health personnel. Another possible study for laboratory research would be a test of the length of time required for Rational Emotive Therapy or behavior modification to be reflected on paper and pencil dependent measures. In this study the rate of behavior change was much higher than attitudinal change. It would be worthwhile to determine if the lack of attitudinal change was a function of time. The last recommendation for clinical research would be to determine how cognitive restructuring Operates to increase or decrease behaviors. One could begin by asking clients to keep a diary of thoughts surrounding particular behaviors. Such a simple effort could point to a direction for future research. For example, it might give an estimate 70 of whether antecedent or consequent thoughts are crucial to particular behaviors. Conclusion Lazarus (1973) highlighted the problems of clinical treatment and research with humans in his discussion counter- ing Wolpe's statement that "Human neuroses are like those of animals in all essential respects" (Wolpe, 1968, p. 559). He challenged Wolpe's statement by saying: When confronted by people intent on self-destruction, torn asunder by conflicting loyalties, crippled by too high a level of aspiration, unhappily married because of false romantic ideals, or beset by feelings of guilt and inferiority on the basis of complex theo- logical beliefs, I fail to appreciate the clinical significance of Wolpe's (1958) neurotic cats and some- times wish that life and therapy were really as simple as he would have us believe (Lazarus, 1973, p. 13). The current study was an attempt to incorporate one human complexity, cognition, in the form of Rational Emotive Therapy, into behavior modification research. This approach is consistent with Lazarus' (1973) view that behavioral techniques can be used as a starting point for increasing therapeutic effectiveness. APPENDICES 71 APPENDIX A QUESTIONNAIRES 72 APPENDIX A Questionnaire 1 (Lawrence, Evaluation by Group_Member Name: 1. 1973) Date: The following are the stated problems and objectives that you worked on in the group. Please rate the extent to which each problem has changed by placing an X in one of the boxes. (1) Problem Objective Very Much Slightly No Slightly Very Much Worse Worse Worse Change Better Better Better (2) Problem Objective Very Much Slightly No Slightly Very Much Worse Worse Worse Change Better Better Better (3) Problem Objective Very Much Slightly No Slightly Very Much Worse Worse Worse Change Better Better Better 73 74 Number of Client Relationships Check which of the following relationships you think have improved in some way since you have been in the group: ___Work supervisorcntemployer ___FriendsOQ a "QSIBOHHOL be HO>OA tiumjebsflufiwao>wg \ oz mow 02 mo» DCOHHO pee bmflaepoeb ceoBDOQ pobeflbomoc hayOSbsE coon mes Oamom OLD be: so MOLLOLB xOOLU bsmebeosb LLHB mmooosm webmafloflbse bmoe .O bcoE Iboosb LOHB mmooosm poboomxo seep OHOE .p mmooosm peoEueOHb we HO>O~ peboomxo .O bcoabeesu LDHB mmooosm eobooaxo Cebu mmOH .Q ssosfla bassosu OEOOMDO UCOEDQOHD OHQQHO>®MCS umOE .6 "m basses “ m meadow "a sesame ”v Oamom "m breads ”m oflmom "m assess um anom "H Damage "H OHeOm mqm>mq EZMZZH