rHESIS ‘JLIBRARY Ill!lllllllllzllljllsllllllllllllslflllllllfllfllfll “5233“ This is to certify that the thesis entitled Client-Therapist Complementarity as it Relates to the Process and Outcome of Psychotherapy presented by Gerald Lee Gaffin has been accepted towards fulfillment of the requirements for . muegreein PSYChOlOgy (Mme Ame. Major professor Date 2.12.1!qu' 0-7 639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS-z Place, n book return to mow circulation recon [,7 a ‘ h b . ""315 1‘ b: r we 713°: .l‘ _‘. ff 1. ; ;n ' “”3" ’ ’ 233 fiyfiw I is“ :5! r3, ‘5 CLIENT-NERAPIST C(MPLEJENTARITY AS IT REIATES TO THE PROCESS AND OUTCOHElOF’PSYCHOTHERAPY By Gerald Lee Gaffin, ILA. A DISSERTATION Su‘mitted to Michigan State University in partial fulfillment of the requiraents for the degree of DOCTOR OF PHILOSOPHY Department of Psycholog 1981 ABSTRACT CLIENT-NWIST CCMPLD‘IENTARITY AS IT RELATES TO THE PROCESS AND CUTCQiE 0F PSYCHOTHERAPY By Gerald Lee Gaffin, M.A. This study investigated the relationship between levels of client-therapist complementarity during three stages of psychotherapy and therapy outcome. Sullivan (1953) refers to behavioral complaen- tarity as an instance in which the needs of one person interact with the needs of another in such a way that both tubers derive satisfac- tion. In toms of the Leary Interpersonal Circunplex (1957). which was used to categorize client and therapist statements in this study. complementarity occurs on the basis of reciprocity on a dominance— suhnissive axis and on the basis of correspondence on a love-hate axis. Eighteen clients seen by 18 therapists at the Michigan State University Psychological Clinic were investigated. Complementarity levels were obtained from the Leary Interpersonal Checklist and from content analysis of therapy tapes. Therapy outcome was assessed using the Symptom Checklist (SCL-90) and the Therapist and Client Post- Therapy Forms. The following hypotheses were investigated. hypothesis}: The level of self-rated client-therapist comple- nentarity at the start of therapy will be positively related to client outcome of psychotherapy. Evidence supporting this "matching” hypothesis was found in the positive correlation between the Therapist Post-Therapy Form and self- rated client-therapist complementarity (r = .4380, p = .035). hypothesis He a b! The level of behaviorally-rated client-therapist complementarity during the course of therapy will be related to client outcome of psychotherapy. The Client Post-Therapy Form exhibited a now-significant but marked tendency for higher levels of overall therapist and client complementarity to be associated with more successful outcome (1‘: 0&317, P: 0072: = ohlfi. P3 00%). Early Stage of Therapy Hmthesis III: During the early stage of therapy. the level of behaviorally-rated therapist complementarity will be positively related to client level of maladjustment. A near significant correlation in the opposite direction of that hypothesized was found between client maladjustment and behaviorally- rated therapist complementarity (r = -.3824, p = .059). hymthesis Na in b: During the early stage of therapy, the level of behaviorally-rated client-therapist complementarity will not be related to client outcome of psychotherapy. This hypothesis was primarily supported. However, contrary evidence was exhibited by a weak positive relationship between the Client Post-Therapy Form and therapist and client complementarity (r: .3208, pg .097: r3 0%“. p8 0080). Middle Stage of Therapy Hypothesis Va is b: During the middle stage of therapy, the level of behaviorally-rated client-therapist complementarity will be negatively related to client outcome of psychotherapy. No significant correlations were obtained. End Stage of Therapy Hmthesis VIa a b: During the end stage of therapy. the level of behaviorally-rated client-therapist cmplementarity will be positively related to client outcome of psychotherapy. The Client Post-Therapy Form significantly correlated in a positive direction with behaviorally-rated therapist and client complementarity (1'8 °h555e P' 00298 1" 038970 P3 '055)0 Complgentarity Patterns in Therapy Hypgthesis VIIa e b: Behaviorally-rated client-therapist comple- mentarity which significantly decreases from the early to middle sessions and significantly increases from the middle to the end sessions will be positively related to client outcome of psycho- therapy. Significant differences in therapist and client complaentarity were not found across therapy stages. Hypgthesis Villa at b: Behaviorally-rated client-therapist comple- mentarity which does not significantly change in level over the three stages of therapy will be negatively related to client outcome of psychotherapy. As a result of the lack of change in therapist and client comple- mentarity across stages for all clients. this hypothesis became untenable. Ex lorato H th sis Hypothesis 1: The lessened ability of therapists to relate to lower class clients will result in no significant change in levels of behaviorally-rated therapist cmplaentarity over the three stages of therapy . and consequently negatively relate to client outcome of psychotherapy. This hypothesis was unsupported. Post Hoe AnalysisuSex Diffggnces Complaentarity levels in like—sex and opposite-sex dyads were examined. No sex differences were obtained. ACKNOWLEDGmENTS In carrying out any research project. it is difficult to thank all those that have been so helpful. Therefore. I am left to thank those most prominent while tendering a silent ”thank you" to the rest. To Dr. Noman Abeles, chairperson of this thesis committee, I would like to express my utmost gratitude. His willingness to give considerably of his time in fostering my personal research ideas as well as share his own insights was greatly appreciated. He helped make research an enjoyable and intriguing part of my training. I would also like to thank Dr. Donald Grumman. Dr. Ralph Levine. and Dr. Dozier Thornton for their encouragaent and helpful ideas in the planning and the execution of this research. Further thanks need to go to the therapists and clients of the Michigan State University Psychological Clinic. Their courage in allowing for a therapy research library to be established there made this study possible. Thanks are also due to the State of Michigan for kindly providing funds for this study. Then once again. special thanks go to Polly Crisp arrl Miquela Rivera for their die-hard assistance in content analyzing this research. Finally. I would like to thank my parents and sister. Living in a family of three M.S.H.'s could do no less than inspire me in my research to unravel the canplexities of the therapy relationship. 11 TABLE OF LIST OF TABLES . . . . . . . . LISTOFFIGURES INTRODUCTION . . . . . . . . . The Interactional Approach REVIEW OF THE LITERATURE . . . Behavioral Analysis System Complementarity . . . . . Client-Therapist Complementarity as to Stages of Therapy Early Stage . . . Middle Stage . . End Stage . . . . EXPERIMENTAL HYPOTHESES . . . Exploratory Hypothesis . "MODS O O O O O O O O O O 0 Subjects . . . . . . Selection of Cases . Measures . . . . . . Selection of Sessions . Behavioral Analysis System Level of Client Maladjustmen Complementarity . . . Therapeutic Outcome . . . REILTS I O O 0 O O O O I O O Complementarity Index . . EXperimental Hypotheses Exploratory Hypothesis . fleece. CONTENTS iii Behavioral Complementarity—-Research Findings. . Client-Therapist Complementarity and Therapeutic mtcme O O O I O I O I O O I O I O O O O 0 O 0.800... 5' it Relates ‘5 00.9.00... 0: acct-emcee Page DlstSION O O O I O O O O O O O O O O O 0 SUMMARY hrlyStageofTherapy . . . . . . Middle Stage of Therapy . . . . . . End Stage of Therapy . . . . . Complementarity Patterns in Therapy Exploratory Hypothesis . . . . . . Post l-loc Analysis--Sex Differences Early Stage of Therapy . . . . Middle Stage Of Therapy e e e e EndStageofTherapy . . . . . . . Complementarity Patterns in Therapy Exploratory Hypothesis . . . . . . Post Hoc Analysis--Sex Differences APPENDICES Appendix A. B. C. D. E. F. LIST OF REFERENCES Client Consent Form . . . . . . . . Therapist Post-Therapy Form . . . . Client Post-Therapy Form Scoring Manual for the Interpersonal Behavior Rating System Summari zed Research Data Individual Therapist and Client Cmplanentarity httems I O O O O O O O O O O O 0 O I O O 0 iv 72 7t} 90 93 97 Table l . 2. 8. 9. 10. LISTG‘TABLES Summary of Therapist Dacgraphic Information . Sum-ary of Client Duographic Information . . Inter-Judge Agreement on Interpersonal BOMVionl Ratings I I I I I I I I I I I I Summary of Significant Correlations and Trends Analysis of Changes in Level of Therapist Complaentarity Over the Three Stages of Then” I I I I I I I I I I I I I I I I I I Analysis of Changes in Level of Client Complaentarity Over the Three Stages of Then” I I I I I I I I I I I I I I I I I I Therapist Complacentarity in Male-Male ani "file-Fable Dyflds (N3 13) e e e e e e e 0 Client Compluentarity in Male-Male and Hale-Fan]: DMB (Na 13 e e e e e e e e Therapist Complementarity in Like-Sex and QEQBItB-ng we (N. 18, e e e e e e e e Client Complementarity in Like-8g and 02281te’sg DMB (N 3 18 e e e e e e e e Page 28 42 67 6? 67 68 Figure l. 2. LIST OF FIGURE The Interpersoml Gimplex e e e e e e e e e e Complementarity Matrix (Dietzel Scoring System) The Interpersoml Cimle e e e e e e e e e e e 0 Changes in Level of Therapist Compluentarity Over the Three Stages of Therapy for the More Successful Outcome Group . . . . . . . . Changes in Level of Client Complementarity Over the Three Stages of Therapy for the More Successful Outcome Group . . . . . . . . Comparison of Therapist Complementarity Levels Over the Three Stages of Therapy for the More Successful Outcome Group for the Dietzel and Abeles (1975) Study and the Present Study . . Comparison of Client Complementarity Levels Over the Three Stages of Therapy for the More Successful Outcome Group for the Dietzel and Abeles (1975) Study and the Present Study vi 47 61 62 INTRODUCTION Typically, psychotherapy research has focused on the input and output dimensions of the therapy endeavor. Research has related client and therapist attributes to the outcome of therapy. From client educational level to therapist orientation, attempts have been made to determine those important ingredients that effect client change. while there have been numerous research studies. the results of the findings are not always clear. Although factors such as client socioeconomic status have borne out as predictors of therapeutic outcome. the exact relationship between these selected client and therapist attributes and outcome is still clouded (Parloff et al., 1978). This lack of clarity is at times attributable to an oversimplification of the therapy endeavor. Therapy is seen as an additive interaction between two people who bring personal attributes, skills, and values to the therapy situa- tion. but the way in which the client and therapist interact over and above these attributes is not always taken into account. Therapy researchers do not often explore the two-way interaction occurring during the therapy hour. Uniformity myths abound (Kiesler, 1966) and clients of similar background and pathOIOgy are equated as are therapists of similar orientations. The result of this is that much therapy research is carried out in which clients and therapists are matched in a research design while the client-therapist interaction is left out as an important factor. It is rare that the therapy "process" is carefully studied scientifically on a dyadic interactional 1 level. While it is true that "everyone is much more simply human than otherwise" (Sullivan. 1953. p. 32), which implies a baselevel similarity that may justify equating individuals, it is also true that "personality is the relatively enduring pattern of recurrent interpersonal situations which characterize a human life" (Sullivan, 1953, p. 111). These are patterns that vary across individuals and which will consequently lead to differences in the therapeutic interaction. It seems imperative to view therapy in this more complex fashion in order to understand it. The Interactional Approach Sullivan is credited as the first to elaborate a theory of personality development based on interpersonal schema. For him, the human infant arrives in this world as an "animal" (Sullivan, 1953, p. 20). Only through interaction with others does the infant become truly human. This would therefore make the infant's early interactions with caregivers very important for the child's interactional style and later personality. The infant is early on a helpless individual and depends on the mothering One to take care of needs. Need tension expressed by the infant's cry evokes a level of anxiety in the care- giver. If this evoked anxiety leads to tenderness by the caregiver. then the infant develops security and anxiety is relieved. If the infant's anxiety meets with anxiety from the caregiver then the infant comes to experience the world as malevolent. The infant does not have the c0gnitive capabilities to reinterpret the world in any other way. In this fashion, the infant's cry and the caregiver's response become a prototype for the infant's later interactions. As the infant matures into childhood, a self-system (Sullivan, 1953) is developed which perpetuates the infant's viewpoint of the world. The child's self- system works at affirming itself while escaping from environmental influences that are not in harmony with itself. Thus the child comes to seek out those interactions that are self affirming and learn that specific behaviors come to elicit responsive behaviors from others. The child will therefore attempt to reinforce others to act in ways that ever increase self-esteem and lessen anxiety. In time, the child's interactional style becomes automatic. The individual's style becomes so ingrained that it is now reflected in the tone of voice. body posture, and verbal communications. Sullivan redescribed psychiatry as the study of interpersonal phenomena (Sullivan, 1953). In his redefinition. he developed the theorem of reciprocal emotion, a theorem which was to have a considerable effect upon the field of psychotherapy. By reciprocal emotion, Sullivan meant that: "Integration in an interpersonal situation is a reciprocal process in which (1) complementary needs are resolved or aggravated: (2) reciprocal patterns of activity are developed or disintegrated; and (3) foresight of satisfaction, or rebuff, of similar needs is facilitated" (Sullivan, 1953. p. 198). Simply stated, interaction patterns are maintained in which the complementary needs of each of the participants are satisfied. Sullivan's works initiated trends apparent today in the current psychological literature aimed at defining interpersonal behavior. Many empirical and theoretical studies have been based on this model which seeks the motivational antecedents of behavior in the inter- personal process rather than as a result of intrapsychic phenomena. The importance of this shift to an interpersonal schema is apparent when one looks at therapy as a relationship between two or more individuals. Studies have focused themselves on those dimensions of the relationship between therapist and client that would elucidate the complex processes of psychotherapy. In viewing psychotherapy in relationship terms, one can see therapy as an extension of all inter- personal relationships in which extended sequences of interaction occur between therapist and client. Through their behaviors, both client and therapist attempt to shape the subsequent behavior of the other (Kell and Mueller, 1966). The result of this perspective is that the outcome of therapy cannot be seen as a unilateral occurrence or as a stagnant process in which client or therapist can be treated as constants. Rather. therapy is a constantly changing, active process. Among the first to investigate empirically the interpersonal process has been the work of Leary (1957). He endeavored to extend Sullivan's initial theorem of reciprocal emotion by proposing that behavior has both eliciting qualities in its ability to pull behavior from others as well as reinforcing qualities in its ability to confirm or discon- firm the behavior of others. This eliciting and reinforcing quality of interpersonal behavior has been further defined by Carson (1969) in his theory of behavioral complementarity. Behavioral complementarity states that particular behaviors on one's part tend to elicit and rein- force other specific types of behaviors in another. The result of this complementarity being in terms of security-maintenance functions for the individuals involved. In order to investigate this complementarity. Leary developed his Circumplex on which an individual's interpersonal behaviors can be plotted. Complementarity is defined as reciprocity between individuals on the dominance-submission axis and correspondence on the love-hate axis. This method of defining sequential interpersonal behaviors has resulted in a considerable amount of evidence in favor of the complementary nature of interpersonal behavior (Heller; Myers, and Kline, 1963; Mueller and Dilling, 1968: Swensen, 1967), and holds promise in mapping out the therapeutic process and making it under- standable. In comprehending the therapeutic process, an innovative feature of the interpersonal system is that it strives to delineate the process whereby the therapist exerts a modifying influence on the client. Governed by the principles of reciprocity and correspondence, the therapist can act so as to elicit and confirm or disconfirm the subsequent behavior of the client. In this manner, the therapist can exert pressure to move the client out of their style of interacting into a more adaptive one. The therapist is able to take a non-complementary stance to client behavior that has been maladaptive. The result of this being a disconfirming of the client's expectation while exerting a pull for more adaptive behavior on the client's part (Carson, 1969; Dietzel and Abeles, 1975). Therapy, though. cannot only be a matter of disconfirming client statements. Therapists that choose this route from the start would be likely to have the client terminate. The therapist and client need to form.a.working relationship (Greenson, 1967) as a basis from which to proceed. The creation of an atmosphere of acceptance and affirmation of each other's interpersonal stance is essential. Accomplishment of this hinges on an understanding and use of the interpersonal system. Within this framework, the therapist initially walks a fine line in therapy in being reciprocating and confirming of client statements deemed adaptive while being accepting but not reinforcing of client maladaptive statements. It is only after the working relationship is established that the therapist can attempt to display non-complementary behavior in an attempt to move the client away from maladaptive patterns. This view of the psychotherapy process parallels what psycho- analysts regard as the "corrective emotional experience" in therapy (Alexander and French, 1946). Through the establishment of a positive transference (the working relationship) in the initial stage of therapy, the therapist is later able to disconfirm.and correct those maladaptive client behaviors that present themselves in the client's transference relationship to the therapist. The therapist in essence displays non- complementary behavior toward the client. One of the main objectives of the present study seeks to relate the level of therapist complementarity during the beginning. middle. and end stages of therapy to the outcome of psychotherapy. A second investigation relates the client's level of complementarity to the eliciting behaviors of the therapist to therapy outcome. Thirdly, an attempt is made at delineating the pattern of complementarity shown between the three stages of therapy as it relates to outcome. Dietzel and Abeles (1975) have found that therapist complementarity displays a U-shaped curve over the course of therapy in instances of successful outcome. Lastly, the relationship between pre-therapy level of client maladjustment and levels of therapist complementarity will be studied. It is believed that clients of greater pathology will have a more limited repertoire of behaviors available to them and will consequently be more invested in trying to limit the therapist to behaviors that are complementary and self—affirming of the client's maladjusted stance. REVIEW OF THE LITERATURE Behavioral Analysis System Based on Sullivan's writings (Sullivan, 1953), it has been theorized that the psychotherapeutic relationship displays an inter- action in which the client and therapist exert reciprocal "pulls" on each other to behave in predictable ways. Freedman, Leary, Ossorio, and Coffey (1951) and Leary (1957) discussed this phenomenon and proposed an empirical method of ordering behaviors on a continuum in a circular fashion, called the Circumplex of Interpersonal Behavior, and labeled these reciprocal interactions. reflexes. Behaviors are described on the circumplex as interpersonally oriented responses and can be plotted on two orthogonal axes labeled dominance-submission and friendly-hostile. Interpersonal behaviors, or mechanisms, are then seen as a blending of the two axes. Interpersonal mechanisms can be rated and categorized in one of four quadrants formed by the dominance-suhnissive and friendly-hostile axes: (1) friendly-dominant, (2) friendly-suhnissive, (3) hostile- su'tmissive, and (1+) hostile-dominant quadrants. Although not used in this study, further division of the quadrants can be made into octants and sixteenths to obtain finer behavioral discriminations. The sixteen categories of the circumplex are: A = Dominate: B = Boast: C = Reject: D = Punish: E = Hate: F = Complain: G = Distrust: H = Camdemn Self; I = Suhnit; J = Admire; K = Trust: L = Cooperate: M = Love; N = Support; 0 = Give: P = Teach (Freedman, 1951; see Figure 1). From this 7 Figure l The Interpersonal Circumplex DOMINANT Hostile Friendly Dominant Dominant Quadrant 0 Quadrant Punish >. u: .4 -’ o ._____.;: 53 m .- C’ a I u Cooperate F Hostile . Condemn 5 Eriendly Subm|SSIve Self . u missuve Quadrant Quadrant SUBMISSIVE circumplex, it can readily be seen that interpersonal mechanisms are a combination of the two axes dimensions. For example, the interpersonal mechanism "Reject" is an equal blending of dominant and hostile mechanisms while the interpersonal mechanism "Cooperate" is a blending of friendly and sutmissive mechanisms, with an emphasis on the friendly mechanism. Another dimension of the Leary system relates to the use of the circumplex as a multilevel approach defining personality data (LaForge et al., l95fi). Three levels of analysis have been delineated. These are: (l) the public level, or those behaviors evidenced by others: (2) the conscious level, or those behaviors evidenced by self; and (3) the private level, or those behaviors tapped by projective techniques. The primary focus of this study lies with the individual behaviors as evidenced by others, the level of public communication. This level of study emphasizes the interactional aspect of behavior as an elicitor of behavior from another. Lastly, this system builds in a dimension of intensity. The distance from the center of the circumplex is a measure of the intensity of the behavior. An individual who is guiding or teaching may be simply managerial or may be autocratic. The closer the rated behavior to the center of the circumplex, the more moderate is the behavioral intensity. Complementarity Sullivan refers to behavioral complementarity as an instance in which the needs of one person interact with the needs of another in such a way that both members of the interaction derive satisfaction. For example, the need of one person to be dominant coincides with the need of another to be submissive in such a way as to result in 10 mutual satisfaction. When complementarity is lessened and need satis- faction is reduced, anxiety is increased. This anxiety motivates the individuals involved to prompt the other to respond to them in security- maintaining ways, that is, in complementary ways. According to the Leary system, interpersonal behaviors are security operations (Sullivan, 1953) aimed at maintaining safety, comfort, and freedom from anxiety. In recognition of this, the purpose of interpersonal behavior becomes the provoking and "pulling” of complementarity from the other. Complementarity occurs on the basis of reciprocity on the dominance-submissive axis and on the basis of corres- pondence on the love-hate axis. This permits the moment-by-moment verbal behaviors of two individuals to be scored on the Leary Circumplex in order to determine behavioral complementarity. For example, a client who emits a friendly-submissive (F-S) response in seeking advice from a therapist would hope to be complemented by a friendly-dominant (F-D) behavior (nurturance, instruction) from the therapist. In this way, the interaction is mutually rewarding to the two involved. The client, through a behavioral stance, has "pulled" a particular behavioral stance from the therapist in which the type of affect exchanged (friendliness with friendliness) from the position requested (dominance following submission) has resulted in complementarity and thus been maximally rewarding. Complementarity may also occur when a hostile- suhnissive (H-S) behavior is followed by a hostile-dominant (Ii-D) behavior. Conversely, interactions that are nonreciprocal on the power dimension (dominance with dominance or su‘unission with su‘unission) and noncorresponding on the affective dimension (love with hate) are maximally noncomplementary. This would apply for hostile-dominant ll behavior followed by friendly-dominance or hostile—submissive behavior followed by friendly-sutmissiveness. The other possible quadrant interactions in which one of the two behavior determinants displays complementarity results in partial complementarity. Either the power dimension is reciprocal or the affect dimension shows correspondence (e.g., H-D with F-S or H-S with H-S). From this view of complementarity of behaviors, Carson (1969) described the outcomes of interactions as having reward or payoff values and showed that each individual's actions and reactions can be weighted for its reward values. Working on this premise, Dietzel and Abeles (1975) constructed a scoring matrix which assesses the degree of complementarity in interpersonal interactions. It is called the Dietzel Scoring System (see Figure 2) and assigns highest weightings to those interactions that display complementarity while assigning the lowest weightings to interactions displaying noncomplementarity. Those interactions that display complementarity on only the power dimension or the affective dimension are assigned intermediate weights. By inserting the proportion of the different types of rated inter- actions into the respective cells and multiplying by cell weights, and summing across the 16 cells, an index of complementarity can be derived for those interactions. Larger complementarity indexes reflect higher levels of complementarity in the interactions whereas smaller indexes reflect less complementarity in the respondent's verbal behaviors to the sender's messages. These complementarity indexes in quantitative form will compose the process variable in the present study. ween 23300 N + 53.30 m + canaoo w + 55.30 m n haugcoauasou £3 £3 £3 a3 ozonafiomtdeoo: £3 £3 23 £3 cinnafismuhfiaofifl £3 £3 £3 £3 pandasonthgcouum 23 £3 £3 £3 oscseeontfienos minnafism agenda—yam unmade—on venison msoufioaoum 33cc: biota Susana 3.38: Armored: wuogunom gossamer waned: avenged—8.3.80 N enema Jaunted Iowans 13 Behavioral Complgmentarity--Resea£gh Findings Research bearing on behavioral complementarity as an important dimension of the therapeutic situation has been repeatedly undertaken with similar results. Heller, Myers, and Kline (1963) studied interview behavior as a function of the client stimulus input. In their investi- gation of the complementarity system they trained 4 actors as clients. Each actor was to display behaviors that were representative of one of the quadrants of the Leary Circumplex. Thus one actor was to display friendly-dominant behaviors, one friendly-submissive behaviors, one hostile-submissive behaviors, and one hostile-dominant behaviors. Each of these actors was presented to 34 therapists-in-training for % hour interviews. The hypothesis was that clients would "pull" specific behaviors from the therapists in line with behavioral complemen- tarity. Results of the study indicated that: l. dominant client behaviors evoked dependent interviewer behavior: 2. dependent client behaviors evoked dominant interviewer behavior: 3. hostile client behaviors evoked hostile interviewer behavior: and 4. friendly client behaviors evoked friendly interviewer behavior. These results suggest that clients may evoke reciprocal behaviors from therapists that are a function of the stimulus qualities of the therapeutic interaction rather than a function only of each individual's personal attributes. Mueller and Dilling (1968) developed a study of which one of the objectives centered on the exploration of the reciprocal behavioral effects of client-therapist interview interactions. Using the interpersonal system of analysis described by Freedman et al. (1951) and Leary (1957), they scored client and therapist statements for ten minute segments of psychotherapy interviews. From these scores, 1h proportions delineating the time spent in each quadrant of the Leary Circumplex were determined. Therapist and client interview behaviors were rank ordered according to the proportion of time spent in the four quadrants and these behaviors were correlated. Results presented support for the interactional "pull" that client and therapist have on each other. Competitive, hostile (H-D) therapist behaviors signifi- cantly and positively correlated with passive. resistant (H-S) client behaviors while supportive, interpretive (F-D) therapist behaviors related positively to client support seeking (F-S) behaviors. An additional study by Mueller (1969) compared client-therapist behavioral complementarity during the initial stage of therapy and in later sessions using the Leary system. Again, it was found that the client and therapist exhibited complementary verbal behaviors. Furthermore, the complementarity increased over sessions. This lent support to the mutual reinforcing qualities of complementary behaviors. Behavioral complementarity has also been found in situations akin to the therapeutic situation. Raush, Dittmann, and Taylor (1959) studied the interpersonal behaviors of 6 hyperaggressive boys in residential treatment. The children were observed twice in 6 life settings and were rated using the Leary Interpersonal System. One of the findings of this study provided evidence for the complementarity hypothesis. Ratings of peer-peer interactions among the boys pointed to passive aggression (H-S) evoking dominant aggression (H-D) and vice- versa. Whereas interactional ratings between the children and adults were mainly composed of friendly-dominant (F-D) responses sent by adults which received friendly-submissive (F-S) responses from the boys. Complementarity was not static over the rating periods but increased 15 with time in accordance with the "pull" of the statements evoked in the interactions. Mackenzie (1968) additionally provided support for the complemen- tarity hypothesis in a study that was not strictly a therapeutic intervention. She rated the interpersonal interactions in both normal and clinic families (father, mother, and son) as they discussed a predetermined topic. Using the Leary system, it was found that normal families exhibited complementary behaviors. These parents' greatest proportion of behaviors fell in the friendly-dominant quadrant while their sons' greatest proportion fell in the friendly-submissive quad- rant. Clinic families also exhibited complementarity, however, it was in different behavioral quadrants. Mother-son interactions were pro- portionately most characterized by hostile-dominant/hostile-suhnissive interactions while father-son interactions were proportionately most characterized by the partially complementary sequence of friendly- dominant/hostile-suhnissive behaviors. In general, it appears that behavioral complementarity is a useful mode of conceptualizing the process of psychotherapy as well as other interactions. Client-therapist interactions do appear to "pull" or evoke specific types of behaviors from the other while having a reinforcing quality toward maintaining behavioral complemen- tarity. The research that has been presented is only a sampling of the research in the area, but is a representative sample. The next step in exploring behavioral complementarity lies in its relationship to the outcome of psychotherapy. l6 Client-Therapist Complementarity and Therapeutic Outcome The ability to provide appropriate complementarity during the therapeutic endeavor so as to effect successful outcome is contingent on the personalities of the client and therapist. Over'many years research has suggested (e.g., McNair, Callahan, and Lorr, 1962; Betz, 1967) that some therapists have greater success with one type of case while other therapists have greater success with another type of case. This observation would suggest that there is some interaction between the therapist's and client's personalities that lead to differ- ential success. In effect, each therapist and client has their unique set of needs and security operations that will define the therapeutic interaction and resulting outcome. One heuristic way of conceptualizing this interaction of personalities is as client-therapist complementarity. The ability of the therapist to provide appropriate verbal behaviors that will "pull" or evoke adaptive complementary behaviors from the client is decisive to therapeutic outcome. This view is taken in a study by Mueller (1969). Mueller found that the client's behaviors with the therapist become increasingly similar to those that occurred in the client's family constellation and with other significant individuals. Con- versely, the therapist's behaviors become more like those verbal behaviors displayed by the client's significant others. The trans- lation of this is that the client and therapist reenact important events in the client's life that have been deterministic of the client's present maladaptive functioning. However, the therapist's verbal behaviors also have a predictable impact on the client along lines of complementarity. The result is that in the reenactment of past 1? conflicts through the personalities of the therapist and client, the therapist has the potential to modify inappropriate client behaviors by taking a stance that "pulls" complementary behaviors from the client that are more adaptive. This same conclusion is voiced by Carson (1969) when he suggests that the role of the therapist is to refrain from responding in a complementary way to client maladjusted behavior while exhibiting complementarity to more adaptive behaviors. In this way, the client is directed into the "therapeutic work" and finally into changed behavior patterns. Realistically then, it becomes crucial to therapeutic outcome that the therapist has a flexible and multifaceted personality (needs and security operations) that allows for the assess- ment and appropriate response to the client's elicitations. Likewise. the client's personality will have a major impact on therapeutic outcome. In most instances, the client's personality is constricted due to past maladaptive interactions. In the therapeutic endeavor, it is hoped that the client can learn new behavioral patterns that are more adaptivé to interpersonal situations. A second factor effecting the ability of the client and therapist to respond appropriately to the other in complementary ways lies in the levels of client and therapist maladjustment. Leary (1957) and Carson (1969) have both noted that maladjustment leads to a constriction of behaviors and an inflexibility in responding to the other. If both client and therapist were to exhibit severe maladjustment, the result would be the inability of either to move about the behavioral quad- rants of complementary behavior. while this inflexibility in the client is to be expected at first, the inflexibility in the therapist results in an inability to provide complementary behaviors to the l8 client or to shift to new appropriate behaviors which would "pull" for*more adaptive behaviors from the client. Furthermore, if the client exhibits severe enough maladjustment and is unable to respond but from a very constricted behavioral stance, then the "pull" on the therapist to respond to the maladjustment in a complementary way is greater. The result of this would be a lessened likelihood of the therapist being able to "pull" the client into a more adaptive behavioral stance and effect change. This difficulty of effecting change in more malad- justed clients will be explored in the present study. It has been postulated that behavioral complementarity is contingent upon the client's and therapist's personalities and levels of maladjustment. However, it is unclear as to the relationship of complementarity to effective outcome. Two differing opinions have been alluded to in the literature. On the one hand, such research as Mueller (1969) or Raush et. al (1959) has shown that behavioral complementarity increases from the beginning to later stages of therapy. The implication of this being that increased complementarity relates to successful outcome. On the other hand, it has been hypothesized that high levels of complementarity are not the "sine qua non" of effective therapy but are rather the end result. Carson (1969) believes that the usefulness of complementarity is in its ability to gradually "push" the client out of maladaptive patterns. It is the job of the therapist to break away from responding in a complementary and reinforcing manner to client statements and to take up alternative behavioral stances that will force the client to respond in new flexible and adaptive ways. 19 In a series of studies by Swensen (1967), the question of level of complementarity was addressed. Swensen believes that client— therapist behavioral complementarity will most likely lead to a satis- fying and harmonious relationship in which certain of the behaviors of the therapist will produce desirable change in their clients. In the first of his studies, he reanalyzed data from a study by Carson and Heine (1962). That study had suggested that there is a curvilinear relationship between client-therapist similarity and psychotherapeutic success in that too little or too much similarity led to reduced effectiveness. In rescoring this data using the Leary Interpersonal Circle, Swensen found that a relationship existed between complemen- tarity and improvement in therapy. In accordance with the curvilinear effect found in the Carson and Heine study, the lowest complementarity and least success was found among the most dissimilar client-therapist pairs. The greatest complementarity and success was found among those pairs in the next to the highest similarity pairings. In order to further substantiate these findings, two additional studies were under- taken. It was felt that the original study was rather imprecise and that it was possibly confounded by the fact that the data was not originally collected to test the complementarity hypothesis. These two further studies measured complementarity prior to therapy through the scoring of client and therapist MMPI protocols using the Leary system. Each client and therapist was located in a quadrant of the interpersonal circle and client-therapist complementarity was assessed. The degree of complementarity was then related to outcome. In the first study it was found that greater client improvement was related to client-therapist complementarity on the dominance-suhnission 20 dimension while no relationship was found for the love-hate dimension. In a replication of this study, complementarity was found to correlate with greater client improvement for both the dominance-suhnission and love-hate dimensions. The relationship between initial complementarity and outcome will be explored in the present study. In all, Swensen (1967) has provided some strong evidence for the hypothesis that greater complementarity is related to greater client improvement. However, as mentioned previously, it has also been postulated by Carson (1969) that it is the job of the therapist to withhold complementarity from the client so as to move the client out of maladaptive ways. This conflict of opinions on the use of comple- mentarity, though, is more apparent than real. Swensen's research is of a pre/post design. This suggests that levels of complementarity at the beginning of therapy are related to therapy outcome. Likewise, much of the cited research on complementarity related levels at the beginning of therapy and at the end. Carson's theoretical stance, on the other hand, speaks to the types of changes that must come about during the course of therapy to allow for successful therapeutic outcome. In essence, these two viewpoints are compatible. Specifically, in successful psychotherapy, client-therapist complementarity is expected to be high in the beginning and end stages while decreasing during the middle or work stage of therapy. Support for this view is found among various theoretical orientations, both analytic and interactional. 21 Client-Therapist Complementarity as it Relates to Stages of Therapy Early Stage The early stage of therapy is marked by the development of the relationship and the building of rapport. In the psychoanalytic framework, this is the stage at which the working relationship is established that allows therapy to take hold (Creenson, 1967). For the interactionalists (Carson, 1969), it is a time when the security operations of the client need to be respected and not reduced for fear of the client terminating prematurely. The therapist therefore spends a sufficient proportion of time engaged in complementary, confirming responses to the client (Swensen, 1967) while at the same time not being so complementary that the therapist overly reinforces maladjusted client behavior. At this point in therapy, the client-therapist complementarity for both successful and less successful clients is similar since this is the relationship building stage in which the therapist endeavors to minimize client anxiety (Cashdan, 1973). Another influence affecting the complementarity offered at this early phase of therapy is the client's level of maladjustment. Clients with severe maladjustment become very constricted in their behavioral repertoires and exhibit a strong stake in securing and maintaining a particular interpersonal stance. Movement from this stance is likely to created marked anxiety. The client may therefore attempt to force the therapist into a complementary stance in excess of that provided to more adjusted clients. Middle Stage The middle stage of therapy is seen as the "work" stage of therapy. This is the point at which Carson's (1969) theoretical 22 views come into prominence. After having established the therapeutic relationship in the early stage of therapy, this is the time when the therapist actively attempts to move the client out of old, inflexible and maladaptive behaviors. Through responding in less complementary ways, the therapist exerts a "pull" on the client to respond using more flexible behavioral repertoires that are of adaptive use to the client. A result of this lowering of therapist complementarity is a loss of security and increased anxiety on the client's part. This initially leads the client to try to respond to the therapist using modes of behavior that have in the past resulted in receiving complementary behavior that was self-affirming. A transference has now been estabe lished in the relationship. It becomes imperative for successful therapeutic outcome that the therapist provide a "corrective emotional experience" for the client (Alexander and French, 1946). This mandates that the therapist respond in ways that do not reinforce the client's old patterns of behavior but allow the client to sample new, potentially adaptive behaviors. With clients who experience less successful out- come, it is believed that the therapist and client become locked in the transference. They continue to confirm each other's non-adaptive modes of behavior through a level of complementarity commensurate with that seen in the early stage of therapy. End Stgge During this stage, the successful client-therapist dyads re- establish a relatively high level of complementarity. If the work of the middle stage has been completed successfully, then the client has learned to use a broader range of behaviors and developed a more 23 flexible, adaptive repertoire. The client has learned to respond to the therapist as a real person and the transference is resolved. In turn, the therapist will respond in a more complementary manner so as to reinforce and validate the client's newly acquired behaviors. In summary, it is believed that more successful clients will show a pattern of complementarity which decreases from the early to the middle sessions and increases from the middle to the final sessions. Less successful clients will tend to show the same level of complementarity throughout therapy. Dietzel and Abeles (1975) found considerable support for this conceptualization of the therapy process and in the main, the present study is a replication of that study using a community population. II. III. IV. VI. HYPOTHESES The level of self-rated client-therapist complementarity at the start of therapy will be positively related to client outcome of psychotherapy. The level of behaviorally-rated client-therapist complementarity during the course of therapy will be related to client outcome of psychotherapy. A. The level of behaviorally-rated therapist complementarity during the course of therapy will be related to client outcome of psychotherapy. B. The level of behaviorally-rated client complementarity during the course of therapy will be related to client outcome of psychotherapy. During the early stage of therapy, the level of behaviorally- rated therapist complementarity will be positively related to client level of maladjustment. During the early stage of therapy, the level of behaviorally- rated client-therapist complementarity will not be related to client outcome of psychotherapy. A. During the early stage of therapy, the level of behaviorally-rated therapist complementarity will not be related with client outcome of psychotherapy. B. During the early stage of therapy, the level of behaviorally-rated client complementarity will not be related to client outcome of psychotherapy. During the middle stage of therapy, the level of behaviorally- rated client—therapist complementarity will be negatively related to client outcome of psychotherapy. A. During the middle stage of therapy, the level of behaviorally-rated therapist complementarity will be negatively related to client outcome of psychotherapy. B. During the middle stage of therapy, the level of behaviorally-rated client complementarity will be negatively related to client outcome of psychotherapy. During the end stage of therapy, the level of behaviorally-rated client-therapist complementarity will be positively related to client outcome of psychotherapy. A. During the end stage of therapy, the level of behaviorally-rated therapist complementarity will be positively related to client outcome of psychotherapy. 2h B. 25 During the end stage of therapy, the level of behaviorally- rated client complementarity will be positively related to client outcome of psychotherapy. VII. Behaviorally-rated client-therapist complementarity which signif- icantly decreases from the early to middle sessions and signifi- cantly increases from the middle to the end sessions will be positively related to client outcome of psychotherapy. A. Behaviorally-rated therapist complementarity which significantly decreases from the early to middle sessions and significantly increases from the middle to the end sessions will be positively related to client outcome of psychotherapy. B. Behaviorally-rated client complementarity which signif- icantly decreases from the early to middle sessions and significantly increases from the middle to the end sessions will be positively related to client outcome of psychotherapy. VIII. Behaviorally-rated client-therapist complementarity which does not significantly change in level over the three stages of therapy will be negatively related to client outcome of psycho- therapy. A. Behaviorally-rated therapist complementarity which does not significantly change in level over the three stages of therapy will be negatively related to client outcome of psychotherapy. B. Behaviorally-rated client complementarity which does not significantly change in level over the three stages of therapy will be negatively related to client outcome of psychotherapy. Exploratoryuflypothesis Research has also pointed to the conclusion that lower socio- economic clients, as defined by levels IV1 and V1 on the Hollingshead and Redlich Index of Social Class (1958), are less likely to obtain satisfaction in therapy (Hollingshead and Redlich, 1958; Myers and Bean, 1968; Strickland and Crowne, 1963). Hollingshead and Redlich primarily attributed this phenomenon to problems arising from dif- ferences in cultural norms, values, and role expectations between 1 The Hollingshead and Redlich Index of Social Class ranges from level I, upper class, to level V, lower class. 26 psychiatrists and their lower SES (socioeconomic status) clients. Further, Nash et al. (1965) found that client "attractiveness" correlated negatively with various social class indices. These findings would seem to point to difficulties in working with lower class clients. Therefore, it is hypothesized that: l. The lessened ability of therapists to relate to lower class clients will result in no significant change in levels of behaviorally-rated therapist complementarity over the three stages of therapy, and consequently negatively relate to client outcome of psychotherapy. METHODS This study was a part of the Michigan State University Psycholog- ical Clinic's Psychotherapy Research Project. The Clinic is a training and research agency of the Department of Psychology and serves as a low cost clinic to adults, children, and families in the greater Lansing Community. Subjects Two groups of subjects provided data for this study-—therapists at the Michigan State University Psychological Clinic and their clients. A. The therapist group was composed of all therapists at the Clinic who consented to take part in the study and who had seen at least one research client in individual psychotherapy for a minimum of 15 sessions. This resulted in a group of 18 therapists. All therapists were in at least their second year of graduate training. Many of the therapists had M.A. degrees or its equivalent (2 or*more years of graduate study). One therapist had the Ph.D. degree. Assignment of clients to therapists was made on the basis of available time and matching of client-therapist schedules as well as on the training needs of the therapists. A summary of therapist character- istics, including number of therapy sessions, is included in Table l. B. The client group was selected from the pool of clients apply- ing for individual adult therapy at the Michigan State University PsychOIOgical Clinic. Clients from the community meeting the 27 28 Table 1 Summary of Therapist Demographic Information Level of Number of Race Gender Training? Sessions Therapist 1 w M 8 #2 Therapist 2 W M 3 27 Therapist 3 B F l 26 Therapist 4 W M l 43 Therapist 5 W M 2 46 Therapist 6 W F l 24 Therapist 7 W M 4 17 Therapist 8 W M 2 l9 Therapist 9 W F 3 36 Therapist 10 W M 1 l6 Therapist 11 w M 1 18 Therapist 12 W M 6 43 Therapist 13 w M l 33 Therapist 14 w M l 15 Therapist 15 W M 11 42 Therapist 16 w M 7 71 Therapist 17 v F 3 34 Therapist 18 v F 2 16 9Levels of Training 1 - beginning practicum (2nd year graduate school) 2 — advanced practicum 3 - first 1--time internship with no prior advanced practicum 4 — first p-time internship with prior advanced practicum 5 — advanced practicum following first %-time internship 6 - second é-time internship with no prior advanced practicum 7 - second E-time internship with prior advanced practicum 8 — advanced practicum following second %—time internship 9 Ph.D. with less than 2 years experience 10 - Ph.D. with between 2 and 5 years experience 11 - Ph.D. with more than 5 years experience 29 following criteria were used in this study. They were to be 18 years of age or older, were voluntary clients, and were functioning well enough to fill out those inventories that were required of them in order to participate. A letter explaining the purpose of the study and eliciting the clients' participation and permission for gathering inventories and tapes (see Appendix A) was given to these clients. It was emphasized that participation in the study would have no effect upon their right to therapy and that all materials would be coded and confidential. Clients who chose not to participate were dropped from the study. Table 2 presents a breakdown on demOgraphic variables for the participating clients. Selection of Cases Out of the universe of therapy cases with complete research data, one case was selected randomly of all therapists who saw a client in extended therapy for at least 15 sessions. This procedure was followed in order to allow time separation between the three stages of therapeutic intervention which were to be analyzed and to allow time for the process variables used in this study to develoP. This procedure was also followed because of the uniqueness in the spacing of taperecordings made of the therapy for this study. Measures A. Therapist Measures--Interpersonal Checklist (Leary, 1957). During the course of the study, each therapist was requested to complete the Interpersonal Checklist on themself. This checklist consists of 128 descriptive words and phrases which the rater uses to describe 30 Table 2 Summary of Client Demographic Information Level of School SES Age Race Gender Completed Index" Client 1 24 v F 17th I Client 2 23 v F 14th III Client 3 22 v F 14th v Client 4 26 w M 16th III Client 5 25 v F 16th III Client 6 29 v F 14th III Client 7 54 w M 20th I Client 8 34 B M 14th IV Client 9 28 w M 18th II Client 10 24 w M 14th III Client ll 26 W M 16th III Client 12 41 w M 16th III Client 13 34 v F 18th II Client 14 31 v F 16th II Client 15 33 v F 16th II Client 16 28 v F 15th III Client 17 26 w M 16th III Client 18 26 w M 16th III aSES - socioeconomic status (ranges from level I, upper class, to level V, lower class) W - White B - Black 31 him or herself. From the resulting list of checked attributes, the person's interpersonal style is located on a circumplex defined by two orthogonally positioned axes: a dominant-submissive axis and an affiliative-disaffiliative (love-hate) axis. B. Client Pre-Therapy Measures 1. Interpersonal Checklist. At the time of the intake interview, all clients were requested to complete this measure on themselves (see part A for a description of measure). 2. Symptom Checklist (SCL-9o, Derogatis, 1977). Clients were also requested to complete this checklist at the time of the intake interview. This measure consists of a list of problems with which people are often faced--problems relating to somatization, obsessive- compulsiveness, interpersonal sensitivity, depression, anxiety, hos- tility, phobic anxiety, paranoid ideation, and psychoticism. This measure consists of 90 statements of problems. The client is instructed to check those statements that are presently a problem for them. C. Outcome Measures 1. Therapist Measures--Therapist Form (see Appendix B). A 23 question therapist form (Strupp et al., 1969, shortened version) was given to therapists at the termination of therapy. This form tapped the therapists' subjective beliefs about the effectiveness of therapy and the change that clients made on their problems. All answers were coded and effectiveness of outcome (therapists' view) was a continuous variable across clients measured as the percentage of satisfaction and change scores on the form. 32 2. Client Measures a. Symptom Checklist (SCL—90). The Symptom Checklist was administered once more at the termination of therapy. Scores on the SCL-9O were looked at from the standpoint of changes in client's total intensity score as a result of therapy. Research by Uhlenhuth and Covi (1969) and Uhlenhuth and Duncan (1968a, 1968b) has shown the SCL-9O to be sensitive to changes in the client as a result of therapy. b. Client Form (see Appendix C). A 55 Question client form (Strupp et al., 1969, shortened version) was given to clients at the termination of therapy. This form tapped the clients' subjective beliefs about the effectiveness of their therapy. Selection ofg§essi9g§ Data in the form of audio tape recordings were selected.frcm cases at three points in therapy to assess the interaction patterns between therapist and client. These stages were defined as the beginning, middle, and end of therapy. The third session tape repre- sented the beginning stage of therapy, two mid-therapy sessions represented the middle stage, and the last tape session represented the ending stage of therapy. In all, four sessions were analyzed per case for a total of 72 sessions. The justification for using the third session tape rather than the first was to avoid any purely information gathering procedures that may take place in the first session. Five tapes of the 72 were unrateable due to poor quality sound tracks or due to the lack of taping of the specific session. In these cases, the next therapy session was recorded. For instance, if the third session tape was unusable, then the fourth session tape would be substituted. 33 In rating the individual tape recordings, each recorded session was divided into five equal sections. Each section consisted of approximately 10 minutes of tape. In this study, the third and fourth fifths of each tape were content analyzed. This was done in order to avoid any "hello" and "goodbye" effects that are typically found at the beginning and at the end of therapy sessions. In addition, research has shown that these middle sections are the most stable for the types of variables that this study measures (Karl and Abeles, 1969). Presentation of tapes for content analysis was by random order. Behavioral Analysis System The method of tape analysis used in this study was one initially developed by Freedman, Leary, Ossorio, and Coffey (1951) and later elaborated by LaForge et a1. (1954), LaForge and Sucek (1955). Leary (1957), and LaForge (1963). It is an interpersonal system of behavioral analysis and has been applied in an array of clinical research settings. by Raush et a1. (1959), Raush et a1. (1960). Swensen (1967), Mackenzie (1968), Mueller and Dilling (1968), Mueller (1969), and Dietzel and Abeles (1975) among others. Scoring utilizing this system of analysis located each response unit (uninterrupted speech) of the client or therapist into one of four quadrants defined by two orthOgonally-positioned axes, a dominant- subnissive axis and an affiliative-disaffiliative (love-hate) axis. LaForge (1963) found that in factor analyzing responses to the Inter- personal Checklist, that these two axes could account adequately for the circumplex reflexes and for the study of relational aspects of the motives to each other. Verbs illustrating each of the quadrants formed 34 by these axes include: (1) dominate, teach, give. support (friendly- dominant): (2) love, cooperate, trust, admire (friendly—submissive): (3) submit, condemn self, distrust, complain (hostile-submissive); and (4) hate, punish, reject, boast (hostile-dominant). (See Figure 3). Figure 3 The Interpersonal Circle Dominant Hostile- Friendly- dominant dominant Hostile g ,a— Friendly Hostile- Friendly- submissive su‘tmi ssive Submissive An important scoring consideration in this method of analysis lies in examining the interaction between the therapist and client. Each may attempt to establish or elicit an emotional state in the interaction so as to provoke a predictable response from the other. The rater scores the person's communication from the standpoint of empathizing with the communicating person from the position of the person to whom the communication is directed (Freedman, 1951). 35 A detailed scoring manual for this system taken from Dietzel (1971) is presented in Appendix D. The 72 therapy tapes were content analyzed by two judges following a period of training on tape recordings not used in the study. Both judges were fourth-year clinical psychOIOgy graduate students and were qualified for'making the types of clinical judgements needed. Both also had previous experience content analyzing therapy tapes. Ratings of the client-therapist communications were made by the raters independently as they listened to the tapes together. No interactions were allowed between the raters while rating except to check on the response number that they were rating at the time. Reliability of ratings was based on 68 of the 72 tapes and was computed using percent agreement (see Table 3). Level of Client Maladjpstment and Therapist Complementarity Prior to the beginning of therapy, all clients who consented to take part in the study were requested to complete the Interpersonal Checklist. From this measure, a vector score in one of the four quadrants of the circumplex was computed. Leary (1957) and Carson (1969) have proposed that the length of this vector which is defined as the distance between the center of the circle and the client's self-rated coordinate (the point of intersection between the client's scores on the dominance-submissive axis and the love-hate axis) is to be considered an index of client maladjustment. Client scores that fall toward the outer edge of the circle are reflective of a rigid, inflexible stance while those that fall toward the center of the circle are reflective of a fluid, flexible style. Scores centered in the smash mean» one no Rfi.:m so panama 36 ram sown can damn pccsmmmw< pudenda see: ems mom on was ecssnson-eaaenor see NR 8 RN oaamnafismhflomor ear some mus some meannsspom-saescaws Ram dmma ma Anna pseudsomnhaucouna pmmmwwww< Hence comawmmfia emohw< mmsoawmm am never sea: a mover hamsanm mo psosmohmm ommpsoohom mom measmmm mwsfipmm Hmhoa>m£mm HmsowammhmucH co pcmsoohw< owesnuampCH m canoe 37 middle of the circle show an equal weighting among the four behavioral quadrants. Therapists were also requested in the course of the study to fill out the Interpersonal Checklist. This was done to determine self-perceived complementarity between therapist and client. Of special interest was the relationship between the clients' maladjustment as measured with the interpersonal system and scores of client maladjustment as obtained from clients' symptom intensity scores on the Symptom Checklist at the start of therapy. To make this comparison, the rank orderings of the self-rated interpersonal scores were correlated with the rank ordered symptom intensity scores of the Symptom Checklist. The rank ordered correlation coefficient between the self-rated interpersonal system and the Symptom Checklist was rho = .3870, p== .56. While this is not significant at the .05 level, this near—significant trend indicates that the Interpersonal Checklist can be used with caution as a measure of client maladjustment. Therapeutic Outcome In their books on psychotherapy change, Garfield and Bergin (1978), Guman and Razin (1977), and Meltzoff and Kornreich (1970) have all emphasized outcome as a multidimensional occurrence. Therapy outcome is not what is measured individually by change scores on an "objective" MMPI type instrument or by a post-therapy client form or a post-therapy therapist form. Rather, outcome is a combination of all three. Consequently, this study puts forth a "tripartite" model of therapy outcome in which a more objective rating scale is combined with the subjective client and therapist rating scales to form a com— posite outcome picture. To this purpose, the Symptom Checklist, the 38 Client Post-Therapy Form, and the Therapist Post-Therapy Form were utilized in this study. RESULTS Complementaripyplndex Complementarity occurs on the basis of reciprocity on the dominance- subnissive axis and on the basis of correspondence on the love-hate axis. This permits the moment-by-moment verbal behaviors of two individuals to be scored on the Leary Circumplex in order to determine behavioral complementarity. Carson (1969) described the outcome of these interactions as having reward or payoff values and showed that each individual's actions and reactions can be weighted for its reward values. Working on this premise, Dietzel and Abeles (1975) developed a scoring matrix (see Figureiz) which assesses the degree of comple— mentarity in interpersonal interactions. This Complementarity Index, as it was called, was used in the present study to obtain quantitative values for the process measure under investigation (client and therapist complementarity levels). The Index was derived by inserting the pro- portions of the different types of rated interactions into the respective cells, multipyling by cell weights, and then summing across the 16 cells. Cell weightings were established to reflect the relative degree of complementarity in interpersonal interactions. Interpersonal theory, as reviewed in the introduction, has put forth the proposition that interactions that maximize complementarity are the most rewarding to those participating and increase the likelihood of further eliciting 39 4o Figure'Z Complementarity Matrix Respondent Behaviors Elicitor Behavior (weighted) n proportions H-D F-D F-o H-S Hostile-Dominant (2)p (l)p (2)12 (3)13 (H-D) Friendly-Dominant (llp (2)? (Blp (2)p (F-D) Friendly-Submissive (2)P (3)p (2)P (l)p (F-S) Mostlle-Sul'nissive (3)? (2)1” (UP . (2)p (H-s) Complfggziarity 2 Column + 2 Column + 2 Column + 2 Column complementary behaviors. Therefore, those interactions indicating the highest complementarity were given the highest weightings (3's). These were interactions that were reciprocal on the power dimension (dominance with submission) and corresponding on the affect dimension (love with love or hate with hate). Conversely, interactions which were nonreciprocal on the power dimension (dominance with dominance or submission with submission) and noncorresponding on the affect dimension (love with hate) were the least complementary and received the lowest weightings (1's). Interactions which exhibited partial complementarity, reciprocal on both dimensions or corresponding on both dimensions, received a middle weighting (2's). Research by Dietzel 41 and Abeles (1975) has shown that while these weightings do not exactly fit the proportions of scores falling in each category, they do reflect the relative proportions of behaviors elicited. Experimentalgfiypotheses: hypothesis I: The level of self-rated client-therapist complementarity at the start of therapy will be positively related to client outcome of psychotherapy. Three separate measures were used in this study to assess client outcome of psychotherapy. These included the clients' Symptom Checklists (30L-90) done pre/post therapy, a post therapy measure of client perceived outcome (Client Post-Therapy Form), and a post therapy measure of therapist perceived client outcome (Therapist Post-Therapy Form). Client-therapist complementarity as assessed from client and therapist self-rated scores on the Interpersonal Checklist did not significantly correlate with client outcome as measured by the Symptom Checklist change scores (r = .0490. p > .10) or by the Client Post-Therapy Form (r'= .0089, p > .10). However, therapists' perceived outcome of therapy (Therapist Post-Therapy Form) did positively correlate with self—rated client-therapist complementarity (r'= .4380, p = .035) (see Table 4). hypothesis II: The level of behaviorally-rated client- therapist complementarity during the course of therapy will be related to client outcome of psychotherapy. The design of this study required separate analyses of client and therapist complementarity. Therefore, the above prediction is tested by two separate hypotheses. 42 Ascohpv Acconpv Ascwnpv Aecmhpv Aecaaee-mv Aesmupv Acoaaoo-mv N00. nmo. mmo. owo. moo. one. ore. mac. mmo. mmmn.: ammm.+ mmm:.+ seem.. momm.+ :mwm.u ends.. sem:.+ omm:.+ d“. mcuoom owsmno pmaaxoono soeashm mmcaemn Suck saswore-onoa oceans wmcapmu Enos sasuore-enoa scones exceeds such sensore-onoe scones mmcaemh Enos saswcre-onoa accuse ososemswemamz mo Hosea accede mwsapmu shes sacsore-emoa accede mmcapmh euoa somwcre-omoe secede escapee Enos mamnoneuemom awfimmmmne m manmfinm> soumodcsa when: annexe mummy umdamwnfip ma mo 2 as so woman woodpsamhmoo asses: eunuchm nonhuman mucoaao Ammmv nspmvm oasosoomoaoom Hozoq hashes» mo owmpm use msahse mHo>oH soawoocososasoo accede eoemw-saaewoaanrom anemone mo owned ecu msahse mao>oa meandvsosoaasou undamamne vopmnnhadmno«>mnom hashes» we dream Adamo mcdhse mHo>mH soawmocosoaoeoo scones eccew-saflswoa>erem hashes» we dream hands mcause mao>ma apauuusosmaasoo unflasnone eopmunhaamuoasmnom hashes» Mo cream hands madame mHo>mH apfiumpcosmaasoo undernona empmnahaasnoa>snom manner» we mwasoo one msfimse mao>ma seawescosoaasoo access eoemw-seamwoa>srom anemone Mo mmhsoo use mafiaso mam>ma apfiampsosmaasoo vmaauuone scammuhHHmHOd>snom Andamzv mao>ma soaaeoeceosseoo enaaswcre-eccaao teamw-caom < oHndde> dramas ecu mcoapmamhuoo pcmoamacwam mo assessm d wand? .m .N .H 1+3 Ila: The level of behaviorally-rated therapist complemen- tarity during the course of therapy will be related to client outcome of psychotherapy. No significant correlations were found between overall behaviorally- rated therapist complementarity and client pre/post Symptom Checklist change scores (r'= -.O408, p >’.25) or with the Therapist Pest-Therapy Form (r = .1460, p > .25). The Client Post-Therapy Form also did not correlate significantly at the .05 level with overall therapist complementarity but did indicate a marked tendency for higher levels of therapist complementarity to be associated with more successful outcome (r = .4317, p = .072) (see Table 4). The trend would have been significant if a l-tailed test of significance had been utilized. lip: The level of behaviorally-rated client complementarity during the course of therapy will be related to client outcome of psychotherapy. Neither Symptom Checklist change scores (r'= -.0384, p :».25) nor the Therapist Post-Therapy Form (r = .1455, p > .25) exhibited a relationship significant at the .05 level with overall behaviorally- rated client complementarity. The Client Post-Therapy Form, however, did show a marked tendency for clients with more successful outcome to exhibit higher levels of overall complementarity (r'= .4156, p'= .086) (see Table 4). Again, this trend would have been significant if a l-tailed test of significance had been utilized. Hypothesis III: During the early stage of therapy, the level of behaviorally-rated therapist complementarity will be positively related to client level of maladjustment. A near significant correlation in the opposite direction of that hypothesized was found between client maladjustment and behaviorally-rated 1+4 therapist complementarity during the early stage of therapy (r = -.3824, p = .059) (see Table 4). This finding points to the tendency for therapists to exhibit higher levels of complementary behavior during the early stage of therapy to those clients least maladjusted. Liypothesisfl: During the early stage of therapy, the level of behaviorally—rated therapist complementarity will not be related to client outcome of psychotherapy. No correlations significant at the .05 level were obtained during the early stage of therapy between behaviorally-rated therapist complementarity and the three measures of client outcome (Symptom Checklist, r = -.0018, p > .10: Client Post-Therapy Form, r = .3208, p = .097: Therapist Post-Therapy Form, r = .1363, p > .10). Of these outcome measures, though, the Client Post-Therapy Form does show a possible, but weak positive relationship with therapist complementarity. Thus, while this hypothesis is in the main upheld, some slight evidence is to the contrary (see Table l+). Hypothesis IVbx During the early stage of therapy, the level of behaviorally-rated client complementarity will not be related to client outcome of psychotherapy. No correlations significant at the .05 level were obtained between behaviorally-rated client complementarity during the early stage of therapy and the three measures of client outcome (Symptom Checklist, r = -.0581, p > .10: Client Post-Therapy Form, r = 01461;, p = .080: Therapist Post-Therapy Form, r = .1960, p > .10). Again, the Client Post-Therapy Form exhibits a possible, but weak positive relationship with client complementarity and gives slight evidence to the contrary for this primarily confirmed hypothesis (see Table it). ’45 hypothesis Va: During the middle stage of therapy, the level of behaviorally rated therapist complementarity will be negatively related to client outcome of psychotherapy. During the middle stage of therapy, no significant correlations were found between behaviorally-rated therapist complementarity and the three measures of outcome (Symptom Checklist, r*= -.0002, p > .10: Client Post-Therapy Form, r = .1897, p > .10; Therapist Post-Therapy Form, r'= .0314, p > .10). This hypothesis is thus not supported by the data. hypothesis Vb: During the middle stage of therapy, the level of behaviorally-rated client complementarity will be nega- tively related to client outcome of psychotherapy. Client outcome of psychotherapy was unrelated to behaviorally- rated client complementarity during the middle stage of therapy as measured by the Symptom Checklist change scores (r = .0122. p > .10), by the Client Post-Therapy Form (1': .2324, p >..10), and by the Therapist Post-Therapy Form (r = .0342, p > .10 ). This hypothesis therefore goes unsupported. Hypothesis VIa: During the end stage of therapy, the level of behaviorally-rated therapist complementarity will be positively related to client outcome of psycho— therapy. The Client Post-Therapy Form significantly correlated in a positive direction with behaviorally-rated therapist complementarity during the end stage of therapy (r'= .4555, p = .029) (see Table 4). No significant correlations were found with the pre/post Symptom Checklist change scores (r’= -.0747, p >».lO) or with the Therapist Post-Therapy Form (r = .1497, p > .10). 46 hypothesis VIb: During the end stage of therapy, the level of behaviorally—rated client complementarity will be positively related to client outcome of psychotherapy. A near significant correlation was found between the Client Post- Therapy Form and the level of behaviorally-rated client complementarity during the end stage of therapy (r = .3897, = .055) (see Table 4). Neither pre/post change scores on the Symptom Checklist (r'= -.0403, p > .10) nor the Therapist Post-Therapy Form (r'= .1111, p > .10) showed a significant correlation. hypothesis VIIa: Behaviorally-rated therapist complementarity which significantly decreases from the early to middle sessions and significantly increases from the middle to the end sessions will be positively related to client outcome of psychotherapy. As Table 5 indicates, significant differences in therapist complementarity were not found across therapy stages. Rather, the major proportion of the variance in therapist complementarity levels was accounted for by variance within the clients themselves. Few clients exhibited the pattern of therapist complementarity as hypoth- esized across stages (see Appendix F for individual therapist complementarity patterns). As a result, no stage differences were obtained for clients with more successful therapy outcome as measured Table 5 Analysis of Changes in Level of Therapist Complementarity Over the Three Stages of Therapy Source df SS HS F p eta2 Stages 2 0.01 .005 0.13 ns. 00.3% Clients 1? 2.06 .121 60.6% Stages x Clients 34 1.33 .039 39.E% Total 53 3-“0 47 by Symptom Checklist change scores, by the Client Post-Therapy Form, or by the Therapist Post-Therapy Form.2 This hypothesis is therefore unsupported by the data. Figure 4 graphically illustrates the lack of significant changes in therapist complementarity levels over the three stages of therapy for each of the outcome measures for the more successful outcome group. Figure 4 Changes in Level of Therapist Complementarity (CL) Over the Three Stages of Therapy for the More Successful Outcome Group S Symptom Checklist C Client Form (0 $551—- —— T Therapist Form E H o S A? C 4‘7” .0 >< b T .H-’ -- C o. T C 3 1‘1 0 .c e: C 5345L_. z: I l I Early Middle End Sessions 2The more successful outcome group as measured by the Symptom Check- list change scores was composed of those clients exhibiting more than 10% change from their initial symptom intensity level. Clients exhibiting less than 10% change or negative change composed the less successful group. On the Client and Therapist Post-Therapy Forms, the more and less successful clients were defined by a mean split. 48 _hypothesis VIIb: Behaviorally-rated client complementarity which significantly decreases from the early to middle sessions and significantly increases from the middle to the end sessions will be positively related to client outcome of psychotherapy. -As indicated by Table 6, significant differences in client complementarity were not found across therapy stages. The majority of clients did not exhibit the pattern of client complementarity hypothesized (see Appendix F for individual client complementarity patterns). Thus. stage differences did not obtain for clients exhib— iting more successful therapy outcome as measured by Symptom Checklist change scores, by the Client Post-Therapy Form, or by the Therapist Post-Therapy Form. This hypothesis therefore goes unsupported by the data. Figure 5 graphically illustrates the lack of significant changes in client complementarity levels over the three stages of therapy for each of the outcome measures for the more successful outcome group. Table 6 Analysis of Changes in Level of Client Complementarity Over the Three Stages of Therapy Source df SS HS F p etaz Stages 2 0.02 .010 0.30 ns. 00.6% Clients 1? 2.15 .126 65.5% Stages x Clients 34 1.11 .033 33.9% Total 53 3.28 hypothesis VIIIa & b: Behaviorally-rated client-therapist complementarity which does not significantly change in level over the three stages of therapy will be negatively related to client outcome of psychotherapy. 49 Figure 5 Changes in Level of Client Complementarity (Cl) Over the Three Stages of Therapy for the More Successful Outcome Group ---'S C "---T Symptom Checklist Client Form Therapist Form Kin Mean Client Cl Values ‘3 '-3 ///c3 c: to TX( 0'30 6f Early Middle End Sessions As a result of the lack of change in client and therapist complementarity across therapy stages for all clients, whether exhibiting more or less successful therapy outcome, Hypotheses VIIIa and b became untenable (see Tables 4 and 5). Exploratory hypothesis ‘hypothesis 1: The lessened ability of therapists to relate to lower class clients will result in no significant change in levels of behaviorally-rated therapist complementarity over the three stages of therapy, and consequently negatively relate to client outcome of psychotherapy. As Table 5 indicated, levels of behaviorally-rated therapist complementarity did not significantly change across therapy stages. As a result, no stage differences were obtained for upper or lower 50 socioeconomic status (S33 clients. Consequently, this hypothesis is unsupported. However, contrary to expectation, lower SE3 clients exhibited a near significant trend toward more successful therapy outcome in this study as measured by pre/post change scores on the Symptom Checklist (r = -.3789, = .062) (see Table 0). DISCUSSION The present study has been an attempt to replicate the findings of Dietzel and Abeles (1975) using a community based population of clients. In their study using college student clients, it was found that levels of client-therapist complementarity exhibited a distinct pattern over the beginning, middle, and end stages of therapy for both successful and nonsuccessful clients. In the early stage of therapy, client-therapist complementarity was at moderately high levels for both successful and unsuccessful clients. This was to promote those relationship building tasks that characterize this stage of therapy. Further, it was found that in this stage of therapy, therapist com- plementarity was directly related to client maladjustment. Those clients more restricted and invested in their maladjusted repertoire evoked more highly complementary responses from therapists. In the middle stage of therapy, client-therapist complementarity was significantly lower for successful therapy clients than for unsuccessful ones. This lower level of complementarity reflects the "working" phase of therapy in which non-complementary, disconfirming behavioral interactions are prerequisite for behavioral change. Finally, it was hypothesized that during the later stages of therapy, the successful therapy clients would have significantly higher levels of client-therapist complementarity since successful clients would have an increased range of newly-acquired behaviors available to them. This was not found to be true. However, as a result of some changes 51 52 in client-therapist complementarity across stages of therapy, global levels of client and therapist complementarity did not differentiate between successful and unsuccessful clients in this study. The present study differed considerably from that of Dietzel and Abeles (1975) and therefore these studies are not totally comparable. While they both studied the relationship between client-therapist complementarity and outcome, they differed in client and therapist populations and in the outcome instruments used. Dietzel and Abeles studied a college undergraduate student population seen by therapists with an average level of training equal to pre-Ph.D. interns (5th year graduate students). The present study, on the other hand, utilized community clients seen by therapists with an average level of training equal to advanced practicum students (3rd year graduate students). Outcome measures were also different in the two studies, although the process measures were the same. However, some of these differences may not be as hmportant as they may appear. Eighty-nine percent of the community clients in the present study were between 22 and 34 years of age and 9M% of them had some level of college training. In some respects, then, this population shared characteristics with a college student population. Similarly, the therapists in the two studies were not totally different. While the therapists in the present study were not as experienced as those in the Dietzel and Abeles study, they were trained therapists and certainly similar to therapists practicing in many community mental health centers. If client-therapist complementarity is a factor in therapy then it should have affected both therapist populations. Finally, the outcome measures were different in the two studies. The Dietzel and Abeles 53 study relied on MMPI scores while the present study utilized Symptom Checklist scores, Client Post-Therapy Form ratings, and Therapist Post-Therapy Form ratings. Although the.MMPI and the Symptom Checklist are similar types of instruments, most of the significant findings in the present study were found on the Client Post-Therapy Form ratings. This problem will be further addressed later in the discussion. In spite of the apparent differences, similarities in process measures used and in certain client and therapist characteristics make a comparison of the present study with the Dietzel and Abeles (1975) study a useful undertaking. As noted, the findings of the present study differed considerably from those found by Dietzel and Abeles. In light of this, and preliminary to reevaluating the hypotheses of Dietzel and Abeles, the original findings of Swensen (1967) were reassessed in Hypothesis I of the present study. This hypothesis investigated the relationship between the pre-therapy levels of client- therapist complementarity and therapy outcome. Swensen had found in separate studies that "more clients improved when client and therapist were opposite on dominance-submission, but on the love-hate dimension greater improvement was found when therapist and client were the same on the love-hate dimensions" (p. 10). In short, this meant that greater client—therapist complementarity led to greater therapeutic success. In part, this finding of Swensen's was upheld in the present study. Three measures of outcome were related to the level of comple- mentarity (match) between therapists and clients on the dominance- suhnission and love-hate axes of Leary's Interpersonal Circle. These measures included pre/post change scores on the Symptom Checklist (SCL-90), a post therapy measure of client perceived success of 54 therapy (Client Post-Therapy Form), and a post therapy measure of therapist perceived client success (Therapist Post-Therapy Form). Of these three measures, the Therapist Post-Therapy Form correlated in a positive direction with optimum levels of client-therapist complemen- tarity. The other two measures exhibited no such relationship. Swensen's (1967) findings, though, are called into question by several methodological considerations. The most serious of these is that he used the MEPI to define the client's interpersonal stance in his studies. Leary and Coffey (1955) have reported only low to moderate correlations between MMPI indices of interpersonal behavior and the actual observed interpersonal behaviors of their patients. Correla- tions of .42 to .47 were found for the dominance—submission axis and .25 to .67 for the love-hate dimension. Furthermore, Swensen's MMPI data on client and therapist interpersonal stance was collected prior to therapy and therefore raises the question of the subject's actual interpersonal stance during the session. This same criticism could be leveled against the pre-therapy matching design undertaken in the present study. As a result of these criticisms in using pre-therapy measures to predict actual client and therapist behaviors during therapy, the second hypothesis of the present study was entertained. It was pre- dicted that the level of behaviorally-rated client-therapist complemen- tarity over the course of therapy would be related to outcome of psychotherapy. This hypothesis, Hypothesis II, was an extension into the therapy process of Swensen's prediction regarding complementarity and outcome and also tested the opposing beliefs of Carson (1969). As mentioned, Swensen believed that greater complementarity would lead 55 to greater improvement in therapy. This he attributed to the assump- tion that complementary relationships were more "harmonious and satis- fying" for both client and therapist and consequently most successful (Swensen, 1967, pp. 7-8). Carson, on the other hand, has presented a hypothesis in opposition to Swensen in which it is postulated that lower levels of therapist complementarity are related to more success- ful outcome. His rationale is that the therapist must avoid responding in a complementary way to the maladjustment and constricted behaviors of the client. Through a non-reinforcing (non-complementary) style, the therapist can direct the client into the therapeutic work and into relinquishing their maladaptive behaviors for new, more useful ones. Although this may threaten the client's security and raise anxiety, Carson believes that this will supply the motive force for change. Dietzel and Abeles (1975) explored the conflicting hypotheses of Swensen and Carson and found that global levels of therapist or client complementarity over the course of therapy were unrelated to outcome. The present study found some evidence to the contrary. In accord with the partial support found in this study for the Swensen matching hypothesis, further support for the relationship between successful outcome and higher levels of complementarity was found in the present study. Of the three outcome measures, client perceived outcome of therapy (Client Post-Therapy Form) exhibited a marked trend in a positive direction toward relating more successful outcome with global levels of client and therapist complementarity. The other two outcome measures showed no relationship. Early Stage of'Therapy Hypothesis.III predicted that during the early stage of therapy, the level of behaviorally-rated therapist complementarity would be positively related with client level of maladjustment. According to interpersonal theory, and especially Carson (1969), it is believed that clients exhibiting greater'maladjustment have a more restricted, or confined repertoire of behaviors available to them. Therefore, they are more strongly invested in maintaining their particular behavioral stance. This is as a result of the extreme anxiety felt by the more maladjusted person when attempting to engage in alter- native behaviors which have become associated with earlier affect laden relationships with significant others. This client will thus attempt to force the therapist into a stance complementary to the narrow, yet comfortable area of functioning for the maladjusted client. Coupled with this, the therapist may be very acutely aware that a reduction in the client's security operations too early in therapy may lead to premature termination. The therapist will then be careful to provide complementary behaviors to their clients, especially the more maladjusted ones whose security is easily threatened. Dietzel and Abeles (1975) investigated this relationship between client maladjustment and therapist complementarity in the early stage of therapy and indeed found that therapist complementarity was greater with the more maladjusted client. The present study though found a trend to the contrary. A near significant causative correlation (p== .059) was found between greater levels of therapist complementarity and greater client adjustment. In defense of this opposing finding, one line of reasoning stands out. As a result of the reinforcing quality 57 of complementary behaviors, therapists may have been very wary about providing much complementarity to the more maladjusted client while being more than willing to provide reinforcement to those areas of strength in the more adjusted client. This is in line with Carson's (1969) general belief that "the therapist must be one person in the client's life--and he will frequently be the only one in a sustained relationship--who does not yield to the client's pressure to supply confirmatory information to the latter's crippled Self" (p. 280). A second hypothesis concerning the early stage of therapy dealt with the level of complementarity exhibited during this time and its relationship with outcome. Dietzel and Abeles (1975) had hypothesized that during the early stage, client and therapist complementarity would be similar for both successful and unsuccessful clients. This hypothesis was based on a theory which sees psychotherapy as undergoing at least three basic stages. First is an early stage marked by relationship and rapport building behaviors. Secondly is a middle stage marked by the "therapeutic work" and thirdly is an end stage marked by integration and increased adjustment. This theory followed in part from a belief that Carson (1969) and Swensen (1967) were both correct in their views of the relationship between client-therapist complenentarity and outcome. It is just that Swensen's findings of greater complementarity correlating with greater outcome and Carson's antithetical belief may deal with different stages of the therapy process. Thus Dietzel and Abeles came to theorize a 3—stage relationship regarding levels of client-therapist complemen- tarity. In the early stage, complementarity would be unrelated to outcome. It was believed that during this stage, complementarity was 58 essential to establishing a therapeutic relationship in‘gll therapy cases. In the middle stage, higher levels of complementarity would be negatively related to outcome since during this "work stage" the therapist would actively try to move the client toward a greater repertoire of behavioral responding. Finally, in the end stage, complementarity would be even more positively elevated in relationship to outcome than in the early stage. This would reflect the client's new, expanded, and more flexible style in responding in a complemen- tary way to the therapist's eliciting behaviors. True to this theory, Dietzel and Abeles found no relationship between levels of complemen- tarity and outcome in the early stage of therapy. Similarly, in the present study, no correlations significant at the .05 level were obtained during the early stage of therapy between client or therapist complementarity and outcome. However, a non—significant but moderate trend was observed between therapist and client complementarity and client perceived outcome of therapy (1': .3208, p2: .097: r’= .3664, I): .080). This is in line with Swensen's research findings and with the earlier findings of the present study concerning the positive trend found between global levels of client-therapist complementarity and outcome. The ramifications of these findings are that stage relationships between complementarity and outcome would all be positive and not changing across stages as found by Dietzel and Abeles. hiddle Stage of Therapy According to the Dietzel and Abeles (1975) theoretical stance, it was believed that client and therapist complementarity at this stage would negatively relate to successful outcome. This would 59 result from this being the "work stage" of therapy during which the therapist actively engages in behaviors designed to pull the client into new modes of responding. If, though, Swensen's views of the positive relationship between complementarity and outcome are accurate as has been found thus far, then the Dietzel and Abeles’ findings should not be obtained in the present study. In fact, a negative relationship was not found in the present study between client or therapist complementarity and outcome. However, neither was a positive relationship obtained as Swensen's (1967) findings would predict. This is perplexing but may reflect some mild support for the Dietzel and Abeles hypothesis while overall continuing to support Swensen's beliefs. It may be that basically throughout therapy, greater levels of complementarity are related to successful outcome. As a result, though, of the "working stage", complementarity is disrupted suf- ficiently enough in the successful cases to nullify the usually positive relationship between complementarity and outcome at this point while not affecting this basic relationship when.measured across all of therapy. End Stage Of Therapy At this stage of therapy, both the Dietzel and Abeles' (1975) theoretical stance and Swensen's (1967) stance would predict greater levels of complementarity positively correlating with success- ful client outcome. For Dietzel and Abeles this correlation would reflect the client's new, more flexible, integrated ability to respond with a greater diversity of complementary behaviors to the therapist's elicitations. For Swensen, this positive relationship 60 would simply give greater weight to the relationship throughout therapy of greater client and therapist complementarity relating positively to successful outcome. In the Dietzel and Abeles study this positive relationship was not found. In the present study, the relationship did hold for the Client Post-Therapy Form. Furthermore, in the present study, the level of therapist and client complementarity did not differ significantly between beginning and end stages of therapy (F = .04, df = 2, 16, ns.; F = .05, df = 2, 16, ns.) when related to client perceived successful outcome of therapy. The implication of this is that the positive relationship between complementarity and outcome found at the end stage in this study rules in Swensen's favor as Opposed to the Dietzel and Abeles investigation. Complementaritngatterns in Therapy Figures 6 and 7 compare the patterns of successful therapy clients for both the present study and the Dietzel and Abeles (1975) study. The differences in patterning are clear. While the Dietzel and Abeles study found some support for a shifting level of client and therapist complementarity across therapy stages, the present study found a fairly constant elevated level of complementarity to be associated with successful outcome. Why the difference? Outside of differences in measures used to assess outcome, one factor stands out between the studies--the client and therapist population. In the Dietzel and Abeles study, the client population consisted of undergraduate students at a university counseling center while their therapists were of an average level of experience equal to pre-Ph.D. interns (5th year graduate students). The clients in the present 61 Figure 6 Comparison of Therapist Complementarity Levels (C1) Over the Three Stages of Therapy for the More Successful Outcome Group for the Dietzel and Abeles (1975) Study and the Present Study D Dietzel and Abeles (1975) G ..__.. Present study (0 ‘1’ _. .355 > E} 3 50 D G 8 0 I: g... E: 845 "'- 2 D Early Middle End Sessions study were all from the community and their therapists were of an average level of experience equal to advanced practicum students (3rd year graduate students). The implications of this are twofold. On the client side, explanations suggest themselves as has been found in the study of YAV183 clients (Goldstein, 1971) or in the study of effects such as Hhitehorn and Betzggjfit' typolOgy. Different thera- peutic methods have been shown by Goldstein to be effective with YAVIS and non-YAVIS clients as have different typologies been 3YAVIS is an acronym which stands forlzoung,‘httractive,.Xerbal, ‘Intelligent, and‘§uccessful. Mean Client Cl Values 62 Figure'? Comparison of Client Complementarity Levels (01) Over the Three Stages of Therapy for the More Successful Outcome Group for thc Dietzel and Abeles (1975) Study and the Present Study D Dietzel and Abeles (1975) 55 G Present study 50;- G D 45 D 40""' Early Middle End Sessions shown to be effective with psychotic and non-psychotic clients. Similarly, college student clients differ from community clients on many dimensions, e.g. age, such that different patterns of client— therapist complementarity may be associated with successful outcome. While it has already been noted that the college and community popula- tions in the two studies had many similarities, they may also have had important differences. A second source of variance lies with the therapists. The therapists in the present study were less experienced therapists as compared to those in the Dietzel and Abeles study. These less experienced therapists may rely somewhat more on relationship 63 variables in providing successful outcome to their clients and may not have the requisite skills to direct their clients through the changing stages of therapy as do the more experienced therapists. This could easily lead to the different patternings of therapy, both associated with successful outcome, seen in the two studies. In fact, Auerbach and Johnson (1977), in a broad review of the research done on therapist experience, conclude that therapist experience is strongly related to the quality of the therapeutic relationship. For example, Strupp (1958) found that less experienced therapists tend to follow the client's lead rather than take the initiative. Such an effect could result in the higher levels of therapist complementarity seen in the present study among less experienced therapists which would in turn reinforce their clients for showing greater complementarity. However, the situation in regard to the relationship between therapist experience and therapy outcome is less clear. It may be that factors such as the "enthusiasm and infectious optimism" of the newer therapist suffices partially for their lowered skill and allows therapy to be successful albeit in a different way (pattern) from the more experienced therapist (Strupp et al., 1969). Finally, it can not be overlooked that the measures of outcome used to evaluate the patternings of client-therapist complementarity were different in the two studies and that this may have effected the patternings obtained. The Dietzel and Abeles (1975) study used pre/post MMPI profiles to assess outcome while the present study relied on the client's subjective perceptions of the success of their therapy (Client Post-Therapy Form) as the significant outcome indicator. Other, more objective measures were used in the present study, but did not 64 significantly correlate with levels of client or therapist comple- mentarity. This is consistent with Bergin and mebert's (1978) conclusion that significant correlations between different outcome criteria occur consistently across studies but not consistently within studies. The use of client perceptions as outcome criteria, though, are confounded. Clients tend to have overall impressions of their therapists' "facilitativeness" (Gurman, 1973) and of the effectiveness of the therapy they receive. They do not tend to discriminate therapist- skill levels. Thus a measure such as the Client Post-Therapy Form may reflect little more than the client's liking for the therapist. In turn, this liking could lead to the very understandable pattern found in this study between higher levels of complementarity and greater outcome as measured by client "liking". In effect, client "liking" equals a therapist who responds in an affirming, "harmonious and satisfying" (Swensen, 1967) way to their clients. _E;plg;atory Hypothegis Research has suggested that lower socioeconomic clients, as defined by levels IV“ and V“ on the Hollingshead and Redlich Index of Social Class (1958), are less likely to obtain satisfaction in therapy (Hollingshead and Redlich, 1958; Myers and Bean, 1968: Strickland and mene, 1963). This led to the exploratory hypothesis that non-changing complementarity over the course of therapy would be associated with lower class clients and as a result lead to less successful therapy outcome. However, complementarity did not “The Hollingshead and Redlich Index of Social Class ranges from level I, upper class, to level V, lower class. 65 significantly change across therapy for any of the clients and this hypothesis was made untenable. It is of interest to note, though, that in this study, lower class clients had a near significant trend toward more successful outcome of therapy as measured by pre/post change scores on the Symptom Checklist. This is contrary to the findings of the vast majority of studies linking lower socioeconomic clients with less significant outcome. In most studies this negative associa- tion is in part the result of premature termination of lower class clients. In the present study, lower class clients, as did all clients, continued until termination. This may possibly have weeded out the vast majority of lower class clients with negative outcomes. Yet, this does not fully negate the fact that lower class clients may make significant change if kept in therapy. Similar results contradicting the absolute negative relationship between lower class clients and outcome have been found by Albronda et a1. (1964) and Frank et a1. (1957) among others. Post Hoc Analysis--Sex Differences A further comparison was made between the Dietzel and Abeles (1975) study and the present one. This dealt with the interaction between the sex of the client and therapist as it effected levels of complementarity. The majority of past research has indicated that client and therapist gender is unrelated to process and outcome measures of therapy. Studies of therapist empathy levels (Cartwright and Lerner, 1963), client verbal dependency expressions (Alexander and Abeles, 1968), and ratings of client symptom relief and satisfaction (Sher, 1975), for example, have all indicated non-significant 66 differences for like and opposite-sex dyads. However, it was believed that in considering role expectations for'males and females (assertive— ness versus submission) that levels of complementarity could be effected by the gender match of the client and therapist. To investigate this, Dietzel and Abeles utilized complementarity scores from the middle stage of therapy. It was during this stage that they found a significant differentiation in level of complementarity between successful and unsuccessful clients. In their analysis of client and therapist gender, no significant differences appeared between male-male versus male-female (therapist-client) dyads nor between like sex (male-male; female-female) dyads versus opposite—sex (male-female: female-male) dyads as these related to levels of therapist and client complementarity. In the present study, this question was also investigated utilizing complemen- tarity scores during the end stage of therapy. The end stage was selected because it was this stage that was most representative in the present study of differences in complementarity levels between more and less successful clients. As Tables 7-10 indicate, no significant sex differences were found. These results. which are in keeping with the Dietzel and Abeles (1975) study, lend further support to the hypothesis that levels of complemen- tarity are determined by actual client and therapist verbal interactions and not by other factors occurring in the therapeutic situation. 67 Table 7 Therapist Complementarity in Male-Male and Male-Female Dyads (N = 13) Sex Therapist-Client N Mean SD t df Male-Female 7 2.62 .32 (two-tailed test) Table 8 Client Complementarity in Male-Male and Male-Female Dyads (N = 13) Sex Therapist-Client N Mean SD t df Male-Male 6 2.71 .28 .228 ns. 11 male-Female 7 2 o 62 o 33 ((two—tailed test) Table 9 Therapist Complementarity in Like-Sex and Opposite-Sex Dyads (N = 18) Sex of Therapist-Client N Mean SD t df Male-Male (N = 6) _ A ' Female-Female (N s 2) 8 2.66 .31 .073 ns. 16 Male-Female (N Female-Male (N 7 3)' 10 2.70 .29 (two-tailed test) 68 Table 10 Client Complementarity in Like-Sex and Oppgsite-Sex Dyads (N = 18) Sex of Therapist-Client N Mean SD t df Male-Male (N = 6) 8 2.68 .28 .022 ns. 16 Female-Female (N = 2) Male-Female (N = 7) 10 2.70 .30 Female-Male (N a 3) (two-tailed test) SUMMARY This study investigated the relationship between levels of client- therapist complementarity during three stages of psychotherapy and therapy outcome. Sullivan (1953) refers to behavioral complementarity as an instance in which the needs of one person interact with the needs of another in such a way that both members derive satisfaction. In terms of the Leary Interpersonal Circumplex (1957), which was used to categorize client and therapist statements in this study, complementarity occurs on the basis of reciprocity on a dominance-submissive axis and on the basis of correspondence on a love-hate axis. Eighteen clients seen by 18 therapists at the Michigan State Univer- sity Psychological Clinic were investigated. Complementarity levels were obtained from the Leary Interpersonal Checklist and from content analysis of therapy tapes. Therapy outcome was assessed using the Symptom Checklist (SOL-90) and the Therapist and Client Post-Therapy Forms. The following hypotheses were investigated. Hypothesis I: The level of self-rated client-therapist complementarity at the start of therapy will be positively related to client outcome of psychotherapy. Evidence supporting this "matching" hypothesis was found in the positive Correlation between the Therapist Post-Therapy Form and self- rated client-therapist complementarity (r = .4380, p = .035). Hypothesis IIa.& b: The level of behaviorally—rated client- therapist complementarity during the course of therapy will be related to client outcome of psychotherapy. 69 70 The Client Post—Therapy Form exhibited a non-significant but marked tendency for higher levels of overall therapist and client complementarity to be associated with more successful outcome (r'= .4317, p = .072; r = .4156, p = .086). Early Stage of Therapy hypothesis III: During the early stage of therapy, the level of behaviorally-rated therapist canplementarity will be positively related to client level of maladjustment. A near significant correlation in the opposite direction of that hypothesized was found between client maladjustment and behaviorally- rated therapist complementarity (r'= -.3824, p:= .059). hypothesis IVa & b: During the early stage of therapy, the level of behaviorally-rated client-therapist complementarity will not be related to client outcome of psychotherapy. This hypothesis was primarily supported. However, contrary evi- dence was exhibited by a weak positive relationship between the Client Post-Therapy Form and therapist and client complementarity (r-= .3208, = .097; r = .3164, p = .080). This trend would have been significant at the .05 level if analyzed using a l-tailed test. Middle Stage of Therapy Hypothesis Va & b: During the middle stage of therapy, the level of behaviorally-rated client-therapist complementarity will be negatively related to client outcome of psychotherapy. No significant correlations were obtained. Therefore this hypoth- esis goes unsupported. End Stage of Therapy hypothesis VIa a b: During the end stage of therapy, the level of behaviorally-rated client-therapist complementarity will be positively related to client outcome of psychotherapy. 71 The Client Post-Therapy Form significantly correlated in a positive direction with behaviorally-rated therapist and client complementarity (r: .4555. p= .029: r= .3897. p= .055)- Complementaripy Patterns in Therapy hypothesis VIIa & b: Behaviorally-rated client-therapist comple- mentarity which significantly decreases from the early to middle sessions and significantly increases from the middle to the end sessions will be positively related to client outcome of psycho- therapy. Significant differences in therapist and client complementarity were not found across therapy stages. hypothesis VIIIa & b: Behaviorally—rated client-therapist comple- mentarity which does not significantly change in level over the three stages of therapy will be negatively related to client outcome of psychotherapy. As a result of the lack of change in therapist and client comple- mentarity across stages for all clients, this hypothesis became untenable. Exploratogy Hypgthesis hypothesis 1: The lessened ability of therapists to relate to lower class clients will result in no significant change in levels of behaviorally-rated therapist complementarity over the three stages of therapy, and consequently negatively relate to client outcome of psychotherapy. This hypothesis was unsupported. However, contrary to expectation, lower SES clients exhibited a near significant trend toward more successful outcome as measured by pre/post change scores on the Symptom Checklist (r = -.3789, p = .062). Post Hoc Analysis-~Sex Differences Complementarity levels in like-sex and opposite-sex dyads were examined. No sex differences were obtained. APPENDICES APPENDIX A CLIENT CONSENT FORM APPENDIX A CLIENT CONSENT FORM Dear Client: The clinic is conducting an evaluation to assess the helpfulness of the services offered here in meeting the needs of our clients. We expect that through this evaluation we will be able to find ways to better serve you. In order to carry out this evaluation, we request your assistance. We will ask you to fill out one or two questionnaires during your initial intake interview, after your last therapy session and sometime after your therapy has ended. In addition, we would like to tape record occasional therapy sessions. These questionnaires and tapes will help us understand your reasons for coming to the clinic and how useful therapy has been for you. All questionnaires and tapes will be held in strict confidence and you will remain completely anonymous. Your right to therapy will not be affected by your decision on whether or not to participate in the evaluation. You also have the right to drop out of the evaluation at any time. If you are willing to participate in this research, please sign the statement below. Sincerely yours, The Staff of the Psychological Clinic H ********************************************H*H***** *****h****7‘3* *-é*~.+* 72 73 I hereby agree to take part in this evaluation research and grant permission for some of my/my child's therapy sessions to be tape recorded. I grant this permission with the understanding that names, questionnaires and recorded materials will be held in strict confidence. APPENDIX B THERAPIST POST-THERAPY mm APPENDIX B THERAPIST POST-THERAPY FORM Please rate each of the following items, comparing the client with other clients whom you see in psychotherapy using the following scale: 1 - very little 3 - some 5 - moderate 7 - fairly great 9 - very great 1. Defensiveness Before 1 3 5 7 After 1 3 5 7 2. Anxiety Before 1 3 5 7 After 1 3 5 7 3. Ego Strength Before 1 3 5 7 After 1 3 5 7 4. Degree of Before 1 3 5 7 disturbance After 1 3 5 7 5. Capacity for Before 1 3 5 7 insight After 1 3 5 7 6. Over-all Before 1 3 5 7 adjustment After 1 3 5 7 7. Personal liking Before 1 3 5 7 for patient After 1 3 5 7 8. Motivation for Before 1 3 5 7 therapy After 1 3 5 7 9. Improvement ex c- Before 1 3 5 7 ted (prognosis)e After 1 3 5 7 10. Degree to which Before 1 3 5 7 countertransference After 1 3 5 7 was a problem in therapy 74 WW ‘0“) 0‘0 0‘0 \O\O \O\O \O\O W0 00 ‘0‘0 75 Continue rating your client in comparison to other clients on the following scale: 11. 12. 13. l4. l5. l6. 17. 18. 19. Degree to which you usually enjoy working with this kind of patient in psycho- therapy Degree of sympto- matic improvement Degree of change in basic personality structure Degree to which you felt warmly toward the patient How much of an "emo- tional investment" did you have in this patient? Degree to which you think the patient felt warmly toward you Over-all success of therapy How would you characterize your working relation- ship with this patient? 1 - very little 3 - some 5 — moderate 7 - fairly great 9 - very great Before 1 3 5 7 After 1 3 5 7 After 1 3 5 7 After 1 3 5 7 Before 1 3 5 7 After 1 3 5 7 Before 1 3 5 7 After 1 3 5 7 Before 1 3 5 7 After 1 3 5 7 Before 1 3 5 7 After 1 3 5 7 1 r extremely poor 3 - fairly poor 5 - neither good nor poor 7 - fairly good 9 - extremely good How satisfied do you think the patient was with the results of his therapy? 1 - extremely dissatisfied 3 - fairly dissatisfied 5 - neither satisfied nor dissatisfied 7 - fairly satisfied 9 - extremely satisfied 76 20. How would you characterize the form of psychotherapy you conducted with this patient? 1 3 5 7 9 Largely supportive Intensive analytical 21. Do you recall any strikingly pleasant experiences that you had during the therapy sessions with this patient? If yes, please mark the number that best indicates the degree of pleasantness. 1 2 3 4 5 6 7 8 9 Mildly pleasant Extremely pleasant 22. Do you recall any strikingly unpleasant experiences you had with this patient? If yes, please mark the number that best indicates the degree of unpleasantness. l 2 3 4 5 6 7 8 9 Mildly unpleasant Extremely unpleasant 23. Over—all, how would you characterize your experiences with this patient? 1 2 3 4 5 6 7 8 9 Unpleasant Pleasant (Questions 1-6, 12, 13, 17, 18 and 23 were used to compute the Thera- pists' ratings on the Therapist Post-Therapy Form in this study.) APPENDIX C CLIENT POST-THERAPY FORM APPENDIX C CLIENT POST-THERAPY FORM For each item choose the answer which you feel best describes your therapy experience. Then circle the appropriate number. EXAMPLE: '1 a. How helpful was therapy for you? 1 - no help 7 - considerable help 3 - little help 9 - very great help 5 - some help By circling the number "7", the person in this example showed that therapy was considerably helpful. 1. How much in need of further therapy do you feel now? 1 - no need at all 7 - considerable need 3 - slight need 9 — very great need 5 - could use more 2. What led to the termination of your therapy? 1 - my decision 5 - mutual agreement 3 - my therapist's decision 7 - external factors (describe briefly) 3. How much have you benefited from your therapy? 1 - a great deal 7 - very little 3 - a fair'amount 9 - not at all 5 - to some extent 4. Everything considered, how satisfied are you with the results of your psychotherapy experience? 1 - extremely dissatisfied 5 - moderately satisfied 2 - moderately dissatisfied 6 - highly satisfied 3 - fairly dissatisfied 7 — extremely satisfied 4 - fairly satisfied 77 10. ll. 12. (a) 78 What impression did you have of your therapist's level of experience? 1 - extremely inexperienced 4 - fairly experienced 2 - rather inexperienced 5 - highly experienced 3 - somewhat experienced 6 - exceptionally experienced How well did you feel you were getting along before therapy? 1 - very well 4 - fairly poorly 2 - fairly well 5 - very poorly 3 - neither well nor poorly 6 - extremely poorly How long before entering therapy did you feel in need of pro- fessional help? 1 - less than 1 year 4 - 5-10 years 2 - 1-2 years 5 - 11-15 years 3 - 3-4 years 6 - 16-20 years How severely disturbed did you consider yourself at the beginning of your therapy? 1 - extremely disturbed 7 - somewhat disturbed 3 - very much disturbed 9 - very slightly disturbed 5 - moderately disturbed How much anxiety did you feel at the time you started therapy? 1 - a tremendous amount 7 - very little 3 - a great deal 9 - none at all 5 - a fair amount How great was the internal "pressure" to do something about these problems when you entered psychotherapy? l - extremely great 4 - relatively small 2 - very great 5 - very small 3 - fairly great 6 - extremely small How much do you feel you have changed as a result of psychotherapy? l - a great deal 4 - very little 2 - a fair amount 5 — not at all 3 - somewhat How much of this change do you feel has been apparent to others? People closest to you (husband, wife, etc.) 1 - a great deal 4 - very little 2 - a fair amount 5 - not at all 3 - somewhat 12. (b) (0) 13. 14. 16. 17. 79 How much of this change do you feel has been apparent to others? Close friends 1 - a great deal 4 - very little 2 - a fair amount 5 - not at all 3 - somewhat Co-workers, acquaintances, etc. 1 - a great deal 4 - very little 2 - a fair amount 5 - not at all 3 — somewhat On the whole, how well do you feel you are getting along now? 1 - extremely well 5 - fairly poorly 2 - very well 6 - very poorly 3 - fairly well 7 - extremely poorly 4 - neither well nor poorly How adequately do you feel you are dealing with any present problems? 1 - very adequately 4 - somewhat inadequately 2 - fairly adequately 5 - very inadequately 3 - neither adequately nor inadequately To what extent have your complaints or symptoms that brought you to therapy changed as a result of treatment? 1 - completely disappeared 4 - somewhat improved 2 - very greatly improved 5 - not at all improved 3 - considerably improved 6 - got worse How soon after entering therapy did you feel any marked change? weeks of therapy (approximately) How strongly would you recommend psychotherapy to a close friend with emotional problems? would strongly recommend it would mildly recommend it would recommend it but with some reservations - would not recommend it 5 - would advise against it CWgtvta I Please indicate to what extent each of the following statements describes your therapy experience. Disregard that at one point or another in therapy you may have felt differently. Use the following code and circle your answer. +2 - strongly agree +1 - mildly agree 0 - undecided -1 - mildly disagree -2 - strongly disagree 18. 19. 20. 21. 22. 24. 25. 26. 27. 28. 29. 30. 80 My therapy was an intensely emotional experience. +2 +1 0 -l -2 My therapy was often a rather painful experience. +2 +1 0 -l -2 I remember very little about the details of my psychotherapeutic work. +2 +1 0 -1 -2 My therapist almost never used technical terms. +2 +1 0 -1 -2 On the whole I experienced very little feeling in the course of therapy. +2 +1 0 -l -2 There were times when I experienced intense anger toward my therapist. +2 +1 0 -1 -2 I feel the therapist was rather active most of the time. +2 +1 0 -l -2 I am convinced that the therapist respected me as a person. +2 +1 0 -l -2 I feel the therapist was genuinely interested in helping me. +2 +1 0 -l -2 I often felt I was "just another patient". +2 +1 0 -l -2 The therapist was always keenly attentive to what I had to say. +2 +1 0 -l -2 The therapist often used very abstract language. +2 +1 0 -1 -2 He very rarely engaged in small talk. +2 +1 0 -1 -2 81 Continue to indicate to what extent each of the following statements describes your therapy experience. Disregard that at one point or another in therapy you may have felt differently. Use the following code and circle your answer. +2 - strongly agree +1 - mildly agree 0 - undediced -l - mildly disagree -2 - strongly disagree 31. The therapist tended to be rather stiff and formal. +2 +1 0 -1 -2 32. The therapist's manner was quite natural and unstudied. +2 +1 0 -l -2 33. I feel that he often didn't understand my feelings. +2 +1 0 -l -2 34. I feel he was extremely passive. +2 +1 0 -1 -2 35. His general attitude was rather cold and distant. +2 +1 0 -l -2 36. I often had the feeling that he talked too much. +2 +1 0 -l -2 37. I was never sure whether the therapist thought I was a worthwhile person 0 +2 +1 0 -1 -2 38. I had a feeling of absolute trust in the therapist's integrity as a person. +2 +1 0 -l -2 39. I felt there usually was a good deal of warmth.in the way he talked to me. +2 +1 0 -1 -2 40. The tone of his statements tended to be rather cold. +2 +1 0 -l -2 82 Continue to indicate to what extent each of the following statements describes your therapy experience. Disregard that at one point or another in therapy you may have felt differently. Use the following code and circle your answer. 41. 42. 1:3. 1:9. 500 +2 - strongly agree +1 - mildly agree 0-tudediced -1 - mildly disagree -2 - strongly disagree The tone of his statements tended to be rather neutral. +2 +1 0 -1 -2 I was never given any instructions or advice on how to conduct my life. +2 +1 0 -1 -2 The therapist often talked about psychoanalytic theory in.my sessions. +2 +1 0 -1 -2 A major emphasis in treatment was upon my attitudes and feelings about the therapist. +2 +1 0 -1 -2 A.major'emphasis in treatment was upon my relationships with people in my current life. +2 +1 0 -1 -2 A major emphasis in treatment was upon childhood experiences. +2 +1 0 -l -2 A major emphasis in treatment was upon gestures, silences, shifts in my tone of voice and bodily movements. +2 +1 0 -l '2 I was almost never given any reassurances by the therapist. +2 +1 0 -l -2 My therapist showed very little interest in my dreams and fantasies. +2 +1 0 -l -2 I usually felt I was fully accepted by the therapist. +2 +1 0 -l -2 83 Continue to indicate to what extent each of the following statements describes your therapy experience. Disregard that at one point or another in therapy you may have felt differently. Use the following code and circle your answer. +2 - strongly agree +1 - mildly agree 0 - undecided -1 - mildly disagree -2 - strongly disagree 51. I never had the slightest doubt about the therapist's interest in helping me. +2 +1 0 -l '2 52. I was often uncertain about the therapist's real feelings toward me. +2 +1 0 -1 -2 53. The therapist's manner of speaking seemed rather formal, +2 +1 0 -l -2 54. I feel the emotional experience of therapy was much more important in producing change than intellectual understanding of my problems. +2 +1 0 -l -2 55. My therapist stressed intellectual understanding as much as emotional experiencing, +2 +1 0 -1 -2 (Questions 1. 3-5. 11, 13-15, 17, 25-40, and 50-53 were used to compute the Clients' ratings on the Client Post-Therapy Form in this study) APPENDIX D SCORING MANUAL FOR THE INTERPERSONAL BEHAVIOR RATING SYSTEM APPENDIX D SCORING MANUAL FOR THE INTERPERSONAL BEHAVIOR RATING SYSTEM General Considerations The interpersonal circumplex, as it will be used in the present study, consists of four categories or quadrants into which all inter- personal behaviors may be rated. The four quadrants are defined by two orthogonal axes: a vertical axis representing the dimension of dominance- sutmission, and a horizontal axis for the affiliative-disaffiliative (friendly-hostile) dimension. A behavior is judged into a specific category by making dichotomous decisions on both axes. In addition, descriptive terms and example statements, to be listed subsequently, are available for each category. In rating the responses, several problems arise. One: affect and content (i.e., words) may, or may not be congruent. For example, consider the client statement "I like you". If this statement is gen- uine it would be rated friendly-submissive (love). If it were stated in a sarcastic tone of voice it would be rated hostile-dominant (punish). If it came after an interpretation which the client did not want to deal with it would be rated hostile-submissive (complain). To minimize the above problems, the following rule was established: affect takes precedence over content. Secondly; within a given unit (uninterrupted speech) one or more 1Freedman et al., 1951. 84 85 shifts in feelings (emotional tone) are possible. For example, the client may begin his/her speech with an openly hostile statement (hostile- dominant) and then shift during the same speech to a self-condemning statement (hostile-submissive). Where this occurs, multiple scorings are required. For the above example, the scoring would be as follows: C : H-D ..... H-S Where there are more than two shifts in the same unit, only the initial and terminal behaviors will be rated. The advantage of this procedure is that it permits a separate analysis of client (or therapist) as (l) respondent to the preceding elicitations of the other party (here, the initial response in the sequence is used), and (2) elicitor (stimulus) of subsequent response in the other (here the terminal behavior is considered). Thirdly; in various cases, raters may use different levels of interpretation. To avoid this, interpretations should not go beyond the immediate context. Descriptive terms and example statements for each categoryz. The following abbreviations will be used: therapist = T client = C Friendly-dominant (F-D) Categopy To dominate, teach, give, support. (1) Dominate (direct, command, diagnostic probe, independent behavior). 2Many of the example statements were obtained from J. Crowder, 1970, Appendix C, pp. 110-123. 86 T or C changes subject, begins new t0pic, asks infor- mation-gathering questions, is dominating, bossy. (2) Teach (advise, give opinion, inform). T or C gives opinion, acts as authority on subject, instructs. (3) Give (help, interpret beyond conscious feelings). Example: T: "If you feel uptight next week we could meet twice." or "Your relationship with your girlfriend appears to be similar to the one you had with your iother." (4) Support (sympathize, reflect feelings, reassure, generalize conscious feelings, approve, nurture, therapeutic probe). As a general rule, reflecting feelings, generalizing feelings, therapeutic probes (when rated here) must come after a statement which contained that data that is reflected, generalized, etc. Support and reassur- ance does not have this limitation. FriendLy-Suhnissive (F-S) Category. To love, cooperate, trust, admire. (1) Love (affiliate, identify with). Examples: "I really like you." "I feel close to you." (2) Cooperate (confide, agree, collaborate). C cooperates with T, works on problem, answers questions, elaborates on reflective statements, agrees with. 87 (3) Trust (depend, ask for help). Example: C: "This problem arose which I hope you will help me with ----." (4) Admire (ask opinion, praise). Example: C: "What should I do?" C: "You're the best therapist in the Counseling Center." Hostile-Submissive (H-S) Category. To submit, condemn self, distrust, complain. (1) Su‘tmit (defer, obey). (a) Sulmission is more to avoid confrontation than to accept validity of statement (sometimes follows an argument). (b) Also, when client expresses extreme helplessness without belief that therapist can help. (c) A mere "Yeah" or "I guess so" response when the therapist is attempting to elicit an elaboration or after the therapist has made a statement about something. (2) Condemn self (depressed, withdrawn). C: "I feel worthless." 0: "I'm no good." 1: "If I were a good therapist, you wouldn't have those feelings." (3) Distrust (suspicious, skeptical). (a) T or C expresses skepticism about other person or his statements. A "What?" following a very 88 clear statement. "Maybe." (b) Suspicious about feelings, motives, etc., of other party. Example: "I don't know if you feel that way about me or not." (4) Complain (rebel, nag, sulk, passively resist). (a) Includes defensive maneuvers, angry withdrawals into silence, resistance expressed in passive ways. (b) Silences of 15 seconds or more where the previous response would suggest that the person is feeling hurt or angry. Hostile-Dominant (H-p) Category To hate, punish, reject, boast (1) Hate (attack, disaffiliate). C: "Go to hell." (2) Punish (be sarcastic, threatening). C: "People are going to keep bugging me until I kill myself." (3) Reject (withholding, competing, accusing). (a) C or T rejects (in hostile tone) the previous statement of the other. Example: "No, that's not so." (b) C and T are arguing, competing, accusing openly. (c) C or T refuses a previous directive. (4) Boast (narcissistic, self-stimulating, intellectualizing). (a) Boastful statements. Example: "I got the highest grade on that last exam." (b) Wandering, free-associating, conversation in which the speaker provides his own stimulation. Usually includes "lists" of events from the past week, rambling statements, etc. 89 (c) C or T intellectualizes. Examples: C: "I haven't worked out my Oedipal conflict yet." T: "What is it that's troubling you?" APPENDIX E SUMMARIZED RESEARCH DATA Key 1. 10. 11. 12. 130 APPENDIX E SUMMARIZED RESEARCH DATA Client Socioeconomic Status Raw Scores (range = 11-77) Pro-Therapy Symptom Checklist Scores (range = 1-4) Post-Therapy S ptom Checklist Scores (range = 1-4 Pre/Post-Therapy Symptom Checklist Change Scores (percent change) ”Match” on Client and Therapist Pre-Therapy Interpersonal Checklists (range = 1-3) Client Maladjustment Scores based on vector lengths from the Interpersonal Checklist (range = 0-100) Be nning Stage Therapist Complementarity Scores range = 1-3) Be nning Stage Client Complementarity Scores range = l-3) Middle Stage Therapist Complementarity Scores (range = 1-3) Middle Stage Client Complementarity Scores (range = 1-3) Ehd Stage Therapist Complementarity Scores (range = 1-3) End Stage Client Complementarity Scores (range = 1-3) Client Post-Thera y Form Ratings (range = -84-47 90 Column 10 ll 12 13 Key 14. 15. 16. 91 Therapist Post-Therapy Form Ratings (range = -43-93) Client Gender (male, female) Therapist Gender (male, female) Column 14 15 16 92 a an an as an an we is. 3 mm on m: 9 an mm m: on o.m~ o.sH o.sH o.rm m.nm o.NH o.Hr 0.: o.rm m.mH m.nm o.NN om.m oo.m mw.m H®.N oo.m oo.N Nw.m em.m mn.m mo.m nm.m mr.m mm.m oo.m mm.m ao.m mm.m nn.~ or.~ ma.~ Nn.~ or.m on.m on.~ nr.~ sn.m mr.m mm.m mn.m sn.~ mc.m rm.m rH.N on.m rn.a mm.m 3m.N Hm.N mn.m dw.m mn.m wo.m do.m Hm.m om.N mm.~ au.m mm.m Hm.m mm.m «N.m om.N om.H am.~ oo.n mm.~ oo.m Ha.m ms.m mn.m sr.m oo.n oo.n Hr.m Ho.m rm.m sm.m mo.m om.~ os.~ wN.N ms.~ oo.m no.~ oo.m sr.~ nn.~ mr.m or.~ oo.n em.m Nr.m mr.m mn.m on.~ an.m am.m mm.~ mN.N 0.5 o.nm m.an o.n~ m.ma o.rn o.rm o.sm m.mH m.om n.wm o.aa o.rn n.ne 0.0m m.om o.sn n.m o.N es o.m o.N o.N o.m o.N o.H o.N o.N o.m o.N o.N o.N o.N o.N o.N N.H a.a o.Ho o.nm n.mm w.m oo.H mr.o 5H.H om.o no.6 pa.o an.o rm.o on.n do.o ne.o mo.a Hp.H sm.n mm.o oN.o am.o mr.o on.H ow.a mm.a am.o am.H mm.a as.o ma.o mm.o o~.a ma.a mo.a nm.a as.a mm.~ aa.n om.o pa an an Na na mm n: me on on ad ma an n: rm 5N Na .mH .5H .0H .ma .3H .NH .H unmade APPENDIX F INDIVIDUAL THERAPIST AND CLIENT CCMPLEMENTARITY PATTERNS APPENDIX F INDIVIDUAL CLIENT COMPIEMENTARITY PATTERNS l.._ t.c B M E h.—/ s B M E: 7-_ t.,c B I! E 10._ V t,c B M E spattern associated Checklist 2. ._ 3, __ - /\ :V. s,t s B M .E B M E 5. 6. c B M .E B M E 8. _ 9._ o———"\. \—_. s,c t QB M B M E 11. __ 12. __ \ VA 8 s,t,c B M E B M E with more successful outcome on the Sympto- tpattern associated with more successful outcome on the Therapist Post-Therapy Form cpattern associated with more successful outcome on the Client Post-Therapy Form 95 More Successful Less Successful More Successful Less Successful More Successful Less Successful 94 Symptom Checklist U Non-U Pattern Pattern 2 | 9 2 5 Therapist Form 3 | 5 1 I 9 Client Form 3 | 6 1 8 I REFERENCES REFERENCES Albronda, H. F., Dean, R. L., and Starkweather, J. A. Social class and psychotherapy. Archives of General Psychiatry, 1964, .10, 276-283. Alexander, F., and French, T. Psychoanalflic therapy. New York: Ronald Press, 1946. Alexander, J. F. Perspectives of psychotherapy process: Dependency, interpersonal relationships, and sex differences. (Unpublished doctoral dissertation, Michigan State University, 1967.) Alexander, J. F., and Abeles, N. Dependency changes in psychotherapy as related to interpersonal relationships. Journal of Consultigg and Clinical Psycholog, 1968, 3, 685-689. Auerbach, A. H., and Johnson, M. J. Research on the therapist's level of experience. In A. S. Gunan and A. M. Razin (Eds.), Effective pgychotherapy: A handbook of researc . New York: Pergamon Press, 1977, 84-102. Bare, C. E. Relationship of counselor personality and counselor- client personality similarity to selected counseling success criteria. Journal of Counseling Psycholog, 1967, y}, 419-425. Baum, O. E., Felzer, S. B., D'Zmura, T. L., and Shumaker, E. Psychotherapy, dropouts and lower socioeconomic patients. American Journal of Orthogychiatgy, 1966, 16, 629-635. Beery, J. H. Therapists' responses as a function of level of therapist experience and attitude of the patient. Joupgl of Consulting and Clinical Psychology, 1970, 2+, 239-243. Benjamin, L. S. Structural analysis of social behavior. Psycholog- ical Revigw, 1974, _8_l, 392-425. Benjamin, L. 5. Use of structural analysis of social behavior (5:153) and Markov chains to study dyadic interactions. Journal of Abnormal Psychology (in press). Bergin, A. E., and Lambert, M. J. The evaluation of therapeutic outcomes. In s. L. Garfield and A. E. Bergin (£213.), handbook of pgychotherapy and behavior chapge: An empirical analysis, 2nd edition. New York: John Riley 3: Sons, Inc., 1978, 139-189. 98 Bergin, A. E., Hurray, E. J., Truax, C. 3., and Shoben, E. J. The apirical emphasis in psychotherapy: A symposium. Journal of Counseling Psycholog, l963, 19, 244-250. Bergin, A. E., and Suinn, R. 1!. Individual psychotherapy and behavioral therapy. In R. R. Rosenzweig and L. H. Porter ($3.), Annual Review of Pa cholo . Palo Alto: Annual Reviews Inc., “197' 5' .' 'V' o1. g5. sob-5%- Berzins, J. I., and Seiuan, E. Subjective reactions of A and B quasi-therapists to schizoid and neurotic communications: A replication and extension. Journal of Consulting and Clinical Psycholog, 1963, £2, 1090-1091. Betz, B. J. Studies of the therapist's role in the treatment of schizophrenic patients. American Jouml of Psychiatry, 196?, L22: 963-9710 Bordin, E. Research stgtggzles in mychothgrapy. New York: John Wiley a: Sons, Inc., 197 . Carson, R. C. Interaction concefis of arsonality. Chicago: Aldine Publishing Company, 1 9. Carson, R. C., Harden, J ., and Shows, I. E. A-B distinction and behavior in quasi-therapeutic situations. Journal of Consulting Psxc. holog, 1964, 2.8, 426-433. Carson, R. C., and Heine, R. U. Similarity and success in therapeutic dyads. Journal of Consulting Psychology, 1962, _2_6_, 38-43. Cartwright, R. D., and Lerner, B. Ehpathy, need to change, and improvement with psychotherapy. Journal of Consulting Psychology, 1963, _2_z, 138-144. Cartwright, R. D., and Vogel, J. L. A comparison of changes in psycho- neurotic patients during matched periods of therapy and no therapy. Journal of Consulting Psycholog, 1960, _2_4, 121-127. Cashdan, S. Interactional pgzchotherapy. Chicago: Aldine, 1969. Cowen, E. L. Social and community interventions. In P. H. Hussen and n. a. Rosenzweig (ns.). Annual Review of P§ycholog. Palo Alto: Annual Reviews Inc., 1973, Vol. _2_4, 423-472. Crowder, J. E. Transference, transference dissipation, and identifi- cation in successful vs. unsuccessful psychotherapy. (Unpublished doctoral dissertation, Michigan State University, 1970.) Cutler, R. L. Countertransference effects in psychotherapy. Journal of Consultim Psycholog, 1958, 22, 349-356. Demos, G. D., and Zuwaylif, F. H. Characteristics of effective coun- selors. Counselor Education and Sumrvision, 1965-66, 5, 163-165. 99 DerOgatis, L. R. SCL-20. Clinical Psychometrics Research Unit: John Hopkins University School of Medicine, 1977. Dietzel, C. S. Client-therapist complaentarity and therapeutic outcome (Doctoral dissertation, Michigan State University, 1971). Dissertation Abstracts International, 1972, 2, 73053-73063 (University Microfilms No. 72-13, #16). Dietzel, C. 8., and Abeles, N. Client-therapist complementarity and therapeutic outcome. Journal of Counseling Psycholog, 1975, 22, 264-272. Fiske, D. R., Cartwright, D. S., and Kirtner, U. L. Are psycho- therapeutic changes predictable? Journal of Abnormal and Social Psycholog, 1964, 62, 418-426. Fix, A. J. , and Haffke, E. S. Relationship between psychotherapy skills and level of training in a psychiatric residency program. Social Science and Mgicine, 1975, 2, 489-491. Frank, J. D., Gliedman, L. R., Imber, S. D., Nash, E. R., Jr., and Stone, A. R. th patients leave psychotherapy. Archives of Neumloa and Psychiatry, 1957, 22, 283-299. Frayn, D. H. A relationship between rated ability and personality traits in psychotherapists. American Journal of Psychiatry, 1968: 2.2.11: 1232‘12370 Freedman, M. 3., Leary, T. F., Ossorio, A. C., and Coffey, M. S. The interpersonal dimension of personality. Journal of Person- ality, 1951, 29, 143-161. Gaffin, G. L. The impact of the therapist on the process and outcome of psychotherapy. (Unpdblished master's thesis, Michigan State University, 1979-) Camsky, N. R., and Farwell, G. F. Counselor verbal behavior as a function of client hostility. Journal of Counseling Psycholog, 1966, l , 184-190. Garfield, S. L., and Bergin, A. E. Therapeutic conditions and outcome. Journal of Abnomal Psycholng, 1971, 12, 108-114. Garfield, s. L., and Bergin, A. E. (3113.), Handbook of psychotheram and behavior change: An empirical analysis. New York: John Wiley 8: Sons, 1978. Coldstein, A. P. Psychotheragutic attraction. New York: Pergamon Press, 1971. Gottschalk, L. A., and Auerback, A. H. (Eds.), N thods of research in psychotheram. New York: Meridith Publishing Company, 1966. 100 Greenson, R. R. The techni ue and ctice of choanal sis. New York: International University Press, 1E7. Crigg, A. E. Client response to counselors at different levels of experience. Journal of Counseling Psychology, 1961, 8, 217-225. Gunan, A. S. Instability of therapeutic conditions in psychotherapy. Journal of Counseling Psycholog, 1973, £9, 16-24. Gurman, A. s.. and Razin, A. n. (Eds.), Effective psychotherapy: A handbook of research. New York: Pergamon Press, 1977. Keller, R., Myers, R. A., and Kline, L. V. Interviewer behavior as a function of standardized client role. Journal of Consulting Psycholog, 1963, g2, 117-122. Hollingshead, A. 3., and Redlich, F. C. Social class and mental illness: A community study. New York: John Riley c Sons, Inc., 1958. Holt, R. R. (311.), New horizon for psychotheram: Autonomy as a ppgfession. New York: International Universities Press, Inc., 1971, Holt, R. R., and Luborsky, L. Personality attgms of mychiatristg: A study of methods for selectipg residents. New York: Basic Books, Inc., 1958, Vols. 1 a: 2. Karl, N. R. , and Abeles, N. Psychotherapy process as a function of the time segment sampled. Journal of Consplting and Clinical Psycholog, 1969, 33, 207-212. Kell, 3. L., and Mueller, H. J. ImEct and change: A study of coun- selipg relationshig. New York: Appleton-Century-Crofts, 1966. Kanp, D. E. Correlates of the Uhitehorn 3etz A3 Scale in a quasi- therapeutic situation. Journal of Consultipg ngcholog, 1966, 10. 509-5160 Kiesler, D. J. Some myths of psychotherapy research and the search for a paradigm Psychological Bulletin, 1966, 6 , 110-136. Kiesler, D. J. The pgpcgss of psychotherapy: npirical foundations and systems of analysis. Chicago: Aldine Publishing Canpany, 1973 . Kiesler, D. J., Klein, M. R., and Mathieu, P. L. Sampling from the recorded therapy interview: The problem of seaent location. Journal of Consulting Psycholog, 1965, Q, 337-344. Kopplin, D. A. Eliciting responses in client-therapist interaction: A content analysis of initial psychotherapeutic interviews. (Unpublished doctoral dissertation. Michigan State University, 1965.) 101 IaForge, R. Research use of the ICL. Oregon Research Institute: 0R1 Technical Report, 3, No. 4, 1963. LaForge, R., Leary, T. R., Naboisek, M., Coffey, R. S., and Freedman, M. B. The interpersonal dimension of personality: II. An objective study of repression. Journal of Pgrsonality, 1954, Z}: 129'153' LaForge, R., and Suczek, R. The interpersonal dimension of personality: III. An interpersonal checklist. Journal of Personality, 1955, 24, 94-112. Leary, T. Intemrsonal digggosis of Ersonality. New York: Ronald Press, 1957. Leary, T., and Coffey, M. S. The prediction of interpersonal behavior in group psychotherapy. Psychodrama and Group Psychotheram Monogph, 1955, 23. Lerman, H. A study of some effects of the therapist's personality and behavior and of the clients' reactions in psychotherapy. (Unpublished doctoral dissertation, Michigan State University, 1963.) Luborsky, L. Perennial mystery of poor agreement among criteria for psychotherapy outcome. Journal of Consultin and Clinical Psychology-1971, 32, 316-319. Luborsky, L., Chandler, M., Auerbach, A., Cohen, J., and Bachrach, H. Factors influencing the outcome of psychotherapy: A review of quantitative research. Psychologcal Bulletin, 1971, 25, 145-185. MacKenzie, M. M. The interpersonal behavior of normal and clinic family members. (Unpublished doctoral dissertation, Michigan State University, 1968.) Meltzoff, J ., and Kornreich, M. Research in Eychothgrapy. New York: Atherton Press, 1970. McNair, D. M., Callahan, D. M., and Lorr, M. Therapist ”type” and patient response to psychotherapy. Journal of Consultir_:g Psychology, 1962, 2_0, 425-429. McNair, D. M., Lorr, M., and Callahan, D. M. Patient and therapist influences on quitting psychotherapy. Journal of Consultipg Psycholog, 1963, g, 10-17. Mills, D. H. Liki as a therapist variable in the psychotherapeutic interaction. Unpublished doctoral dissertation, Michigan State University, 1961:.) 102 Mills, D. M., and Abeles, N. Counselor needs for affiliation and nurturance as related to liking for clients and counseling process. Journal of Counseling Psycholog, 1965, 2.2, 353-358. Mills, D. M., Chestnut, H. J., and Hartzell, J. P. The needs of counselors: A component analysis. Journal of Counselipg Mintz, J. What is ”success” in psychotherapy? Journal of Abnonal Psycholog, 1972, §_Q, 11-19. Mueller, H. J. Patterns of behavior and their reciprocal impact in the family and in psychotherapy. Journal of Counseling ngchology, 1969, _l_§_. (2. Pt. 2). Mueller, H. J., and Dilling, C. A. Therapist-client interview behavior and personality characteristics of therapists. Journal of Projective Technigues and Pegonality Assessmen , 1968, 2, 281-2880 Mueller, H. J ., and Dilling, C. A. Studying interpersonal thanes in psychotherapy research. J ouygl of Counseling ngcholog, 1969: .Lé: 50'58' Mullen, J. A. An investigation of the variable of liking in therapy: Its relation to the variables of outcome, empathy, and therapist experience. (Unpublished doctoral dissertation, Michigan State University, 1969.) Mullen, J. A., and Abeles, N. Relationship of liking, empathy, and therapist's experience to outcome of psychotherapy. Jouggl of Counseling Psychology, 1971, 28, 39-43. Murphy, U. A., and Lamb, D. H. The effect of training in psychotherapy on therapists' responses to client hostility. Jouml of Community Pchholog, 1973, _1,, 327-330. Myers, J. D., and Auld, F. Some variables related to outcome of psychotherapy. Journal of Clinical Psycholog, 1955, Q, 51-54. Myers, J. M., and Esan, L'. L. A decade later: A follow-up of §ocia1 class and m ntal illness. New York: John Riley a. Sons, Inc., 1968. Nash, E., Moehn-Saric, 3., Battle, C., Stone, A., Imber. S., and Frank, J. Systematic preparation of patients for short-term psychotherapy. II: Relation to characteristics of patient, therapist and the psychotherapeutic process. Journal of Name and Mental Disease, 1965, M, 374-383. 103 Parloff, M. 3., Uaskow, I. E., and Rolfe, 3. E. Research on therapist variables in relation to process and outcome. In S. L. Garfield and A. E. Bergin (Eda). handbook of psychotherapy and behavior chan e: An emflrical analysis. New York: John Riley a Sons, 1978. Perlman, G. Change in "central therapeutic ingredients” of beginning psychotherapists. Psychotherapy: TheoryI Research ‘m P1364109, 1973’ 1—0’ 16.51. Raush, M. L., Dittman, A. T., and Taylor, T. J. The interpersonal behavior of children in residentialtreatment. Journal of AbnoEal anL Social Psychology, 1959, 52, 9-26. Raush, R. L., Far‘man, I., and Llewellyn, L. G. Person, setting, and change in social interaction. Runan Relatignjs, 1960, _1_3, 305-333- Reissman, 0., Cohen, J., and Pearl, A. (Eds.), Lental health of the mor. New York: The Free Press of Glencoe, 1964, 68-75. Rogers, C. R. A theory of therapy, personality, and interpersonal relationships, as developed in the client-centered framework. In S. Koch (Ed), Psycholog: A study of science. New York: McCraw-Hill Book Company, Inc., 1959, Vol. 3, 184—256. Rogers, C. R., Gendlin, E., Kiesler, D., and Truax, C. 3. 1'5: therapeutic relationship and its has}: ' A ptudy 9f psych?- th_e__ra_;py withvgchizophrenicas. Madison, Uisconsin: Univers ty of Uisconsin Press, 1967 . Schaffer, N. D. The impact of therapist personality variables on psychotherapy process. (Unpublished doctoral dissertation, Michigan State University, 1977.) Scher, M. Verbal activity, sex, counselor experience, and success in counseling. Journal of Counseling Psycholog, 1975, 22, 97-1010 Schiller, L. S. The intake interview in psychotherapy: Client- therapist complementarity and role behavior. (Unpublished doctoral dissertation, Michigan State University, 1977.) Schuldt, N. J ., and Smee, P. G. Personal needs of graduate students in psycholOgy. Psychological Repgrts, 1968, 22, 916. Secord, P. E., and 3ackman, c. H. An interpersonal approach to personality. In B. A. Maher (Ed.), Progess_in exmrimentg; research, Vol. 2. New York: Academic Press, 1965, 91-125. Strickland, 3., and Crowne, D. P. Need for approval and the premature termination of psychotherapy. 112111131 of Consulting Psycholog, 1963, 22, 95-101. 104 Strupp, M. The performance of psychiatrists and psychologists in a therapeutic interview. Journal of Clinical Pjsygholog, 1958, 24, 219-226. Strupp, H. H. Psychotherapy: Clinical, researchI and theoretical issues. New York: Jason Aronson, Inc., 1973. Strupp, H. R., Medley, U. M., and Comes-Schwartz, 3. Psychotherapy jpr better or wopse: The probla of neggtive effect . New York: Jason Aronson, Inc., 1977. Strupp, M. R., Lessler, R., and Fox, K. Patients view their gychotherapy. Baltimore: Johns Hopkins Press, 1969. Sullivan, R. S. The intemrsonal theogy of 312mm. New York: Norton, 1953. Sullivan, P. L., Miller, C., and Smelser, U. Factors in length of stay and progress in psychotherapy. Journal gf_Consultipg Psychology, 1958, 22, 1-9. Swensen, C. R. Psychotherapy as a special case of dyadic interaction: Some suggestions for theory and research. Psychotherapy: Theory, ResearchI and Practice, 1967, 4, 7-13. Swenson, C. M. Commitment and the personalit of the successful therapist. In J. Matarazzo et al. (Eds.), Ps chothera . New York: Aldine-Atherton, Inc., 1972, 203-20 . Uhlenhuth, E. M., and Covi, L. Subjective change with the initial interview. Amgrican Journal of Psychotheram, 1969, 23, “15"“29 e Uhlenhuth, E. R., and Duncan, D. 3. Subjective change with medical student therapists: 1. Course of relief in psychoneurotic outpatients. Archives of General Psychiatg, 1968a, .12, 428-438. Uhlenhuth, E. M., and Duncan, D. B. Subjective change with medical student therapists: II. Some deteninants of change in psycho- neurotic outpatients. Archives of Gengral Psychiatg, 1968b, .]'_8: 532-5150. Uaskow, I. E., and Parloff, n. 3. (Eds.), Psychotherapy chapge measures. Rockville, Maryland: National Institute of Menatl Health, 1975. Uhitehorn, J. C., and Betz, 3. A study of psychotherapeutic relationships between physicians and schizophrenic patients. American Journal of Psychiatg, 1954, 512, 321-331. 105 Uhitehorn, J. C. , and Betz, B. A comparison of psychotherapeutic relationships between physicians and schizophrenic patients when insulin is combined with psychotherapy and when psycho- therapy is used alone. American Journal of Psyphia_tg, 1957, 112. 901-910. Uhitehorn, J. C. , and Betz, 3. Further studies of the doctor as a crucial variable in the outcome of treatment with schizophrenic patients. American Journal of ngchiatgy, 1960, 222, 215-223. Hilliams, M. V., Liman, R. S., Rickels, R., Covi, L., Uhlenhuth, E. M., and Mattsson, N. B. Replication of symptom distress factors in anxious neurotic outpatients. Multivariate gmuopl Research, 1968, 3, 199-212. Hinder, C. L., Arnad, F. 2., Bandura, A., and Rau, L. C. Dependency of patients, psychotherapists' responses and aspects of psycho- therapy. Journal of Consultng Psycholog, 1962, 26, 129-134. nICHIGnN STATE UNIV. LIBRARIES mWWW”I”WINWIWIWWIWWWI 31293102558826