PERSPECTIVES ON PSYCHOTHERAPY PROCESS AND EMPA'I'HY A Disserfaflon Ior II“: Degree oI DII. D. MICHIGAN STATE UNIVERSITY Harold S. Steinitz I976 ARY lll3ll”MllfllljfllllllflljlflIII"MIMI WEEQM‘E Ufiveasicy This is to certify that the thesis entitled Perspectives on Psychotherapy and Empathy presented by HAROLD S. STEIN I TZ has been accepted towards fulfillment of the requirements for Psychology 1n Mam MWD Major professor Ph'D' degree ' Date M Mid-Ignaz Sm ' ’ I 7 7-1" ABSTRACT PERSPECTIVES ON PSYCHOTHERAPY PROCESS AND EMPATHY BY Harold S. Steinitz Empathy has been widely studied in order to assess its contribution to the process of psychotherapy. This study's primary purpose was to evaluate the degree to which empathic understanding occurred with two differing client "types." One group, a high subjective discomfort group (elevated MMPI depression and psychasthenia scale scores), has been shown by previous research to be a mean— ingful typology for college students. The second client group, a low subjective discomfort group, consisted of those with significantly lower scores on the MMPI depres— sion and psychasthenia scales. Differences between the two groups regarding age of client, verbal and total scores on a standard college entrance examination, and all other MMPI validity and clinical scales were essen— tially nonsignificant. The completed psychotherapies of 40 college student clients were examined using Carkhuff's (1969) Empathic Understanding in Interpersonal Processes Scale. Harold S. Steinitz The following questions were considered: Do anxious—depressed clients receive more empathy than other clients, do experienced therapists offer more empathy than inexperienced therapists, do therapists increase empathy offered across time, and is there a client sex difference in empathy received? Discussion centered not only on the above, but also on differences in same-sex versus cross—sex dyads in relation to empathy levels generated. While the hypotheses were stated too broadly, the results suggest support for the notion that therapists offer differential amounts of empathy with clients of differing characteristics. The empathy level in any therapeutic encounter depends on the sex of client or therapist, experience level of therapist, and client type. Findings include the following: (a) experienced therapists offer higher empathy in general, but particu— larly with clients who are reporting high subjective dis— comfort; (b) empathy ratings from late sessions of therapy provide far more statistically significant results than do those of early sessions; (c) there is qualified support that anxious-depressed clients receive higher empathy than other clients; (d) female clients as a group do not receive higher empathy than do male clients; (e) same-sex dyads are more likely to achieve high peaks of empathy than are opposite—sex dyads; (f) experienced therapists tend to increase peak empathy Harold S. Steinitz across time; (9) anxious-depressed clients tend to receive higher peak empathy across time, while other clients tend to receive lower peak empathy across time; (h) anxious—depressed male clients receive increases in peak empathy across time, while other male clients receive decreases in peak empathy across time; and (i) senior staff greatly with female clients, but Overall, clients levels of empathy. Peak session seem to generate increase peak empathy across time not with male clients. in this study receive acceptable empathy scores in the late the most significant differences. Low and peak empathy ratings were significantly correlated for early and late sessions, as well as for combined sessions. It appears that therapists offer high empathy levels while avoiding low levels. Implications of these findings suggest the need to control for client personality characteristics even in relatively homogeneous college student client populations. PERSPECTIVES ON PSYCHOTHERAPY PROCESS AND EMPATHY BY Harold S. Steinitz A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1976 ACKNOWLEDGMENTS My deep gratitude is extended to Dr. Norman Abeles. He frequently made himself available for con— sultations and was a particularly important force in my development as a researcher. My respect for him as advisor, chairman, and person has grown as I have come to know him. He treated my work as something valuable and demonstrated serious commitment to his role as chair— man. For all these things, I thank him. To Drs. Albert Aniskiewicz and Griffith Freed, I wish to convey my thanks for reviewing my work, and for graciously accepting committee membership. Dr. Bertram Karon was especially helpful as a statistical consultant during the difficult period of selecting appropriate analysis methods. My raters, Jean Schwartz and Susan Parry, were exceptionally able and understanding. In addition to completing the ratings efficiently, they showed evidence of being rather empathic themselves. To my parents, who instilled in me long ago a sense of joy, purpose, and love for education, I also ii wish to express continued appreciation. They have many times shared in my successes and disappointments, and I know they will celebrate with me now. Finally, I wish to thank my wife, Elaine, for struggling with me, and for furnishing immeasurable amounts of assistance during the course of this research. Every obstacle and every forward-leap were mutually experienced. She provided love, caring, and added incentive which proved to be supportive in the completion of an, at times, arduous task. Empathy, it seems, can be maintained over long periods. iii TABLE OF CONTENTS Chapter Page I. INTRODUCTION AND REVIEW OF LITERATURE . . . 1 Introduction. . . . . . . . . . . 1 Empathy . . . . . . . . . . . . 2 Overview . . . . . . . . . . . 2 Empathy-~Stable Characteristic or Situ- ationally Defined? . . . . . . . 4 Empathy--Criticism and Counter- Arguments . . . . . . . . . . 6 Mixed results . . . . . . . . . 6 validity. O O O O O O O O O O 8 Therapist-Client Interaction . . . . . 12 General Issues . . . . . . . 12 Client Discomfort and/or Disturbance, ‘As Related to Psychotherapy Process and Outcome.' . . . . . . . . . 16 II. EXPERIMENTAL DESIGN . . . . . . . . . 19 Subjects . . . . . . . . . . . . 19 MethOd. O O O 0 O O O O O O O O 26 Hypotheses . . . . . . . . . . . 26 Instruments . . . . . . . . . . . 33 MMPI. O I O O O Q I O O O O O 33 A Scale for Measurement of Empathic Understanding in Interpersonal Processes . . . . . . . . . . 36 III. RESULTS 0 O I O O I O Q 0 O O O O 39 General Considerations . . . . . . . 39 High Versus Low Client Subjective Dis— comfort. . . . . . . . . . . . 41 iv Chapter Client Discomfort Main Effects . . . . Client Subjective Discomfort and Therapist Experience. . . . . . . Summary of Hypothesis 1 . . . . . . . Client Subjective Discomfort and Sex of Client 0 o o o o o o O O 0 Client Subjective Discomfort and Sex of Therapist . . . . . . . . . . Experience of Therapist . . . . . . Experience of Therapist Main Effects . . Therapist Experience and Sex of Client . Therapist Experience and Sex of Therapist . . . . . . . . . . Summary of Hypothesis 2 . . . . . Client Discomfort and Therapist Experience. Empathy Across Time . . . . . . . Empathy Across Time--Client Subjective Discomfort Main Effects. . . . . . Empathy across time--Client discomfort and sex of client . . . . . . . Empathy across time-~C1ient discomfort, therapist experience, and sex of therapist. . . . . . . . . Empathy Across Time—-Experience of Therapist Main Effects . . . . . . Empathy across time--Therapist exper- ience and sex of client . . . - . Empathy across time--Therapist exper- ience and sex of therapist . . . . Empathy Across Time-~Sex of Client Main Effects 0 O O O O O O O O O O Empathy across time—~Sex of client and sex of therapist . . . . . . . Empathy Across Time-~Sex of Therapist Main Effects . . . . . . . Page 41 42 44 45 46 46 48 48 50 52 53 54 54 54 55 56 56 57 57 58 58 Chapter Summary of Hypothesis 4 . . . . . . . Sex of Client . . . . . . . . . . Sex of Client Main Effects . . . . . Sex of Client and Therapist Experience . Sex of Client and Sex of Therapist. . . Summary of Hypothesis 5 . . . . Sex of Therapist Main Effects . . . IV. DISCUSSION . . . . . . . . . . General Considerations . . . . . High Versus Low Client Subjective Dis- comfort (Hypothesis 1).. . . . Client Subjective Discomfort Main Effects . . . . . . . . Client Subjective Discomfort and Exper— ience of Therapist . . . . Client Subjective Discomfort and Sex of Client . . . . . . . . . Client Subjective Discomfort and Sex of Therapist . . . . . . . . . . Experience of Therapist (Hypothesis 2) . . Experience of Therapist Main Effects . . Therapist Experience and Sex of Client . Therapist Experience and Sex of Therapist . . . . . . . . . . Client Subjective Discomfort and Therapist Experience (Hypothesis 3). . . . . . Empathy Across Time (Hypothesis 4) . . . Empathy Across Time—~Client Subjective Discomfort . . . . . . . . . . Empathy across time-~Client subjective discomfort and sex of client . . Empathy Across Time-~Experience of Therapist . . . . . . . . . . Empathy across time——Experience of therapist and sex of client. . . . V1 62 62 63 63 64 65 66 66 66 67 67 68 69 69 70 71 71 Chapter Empathy across time—-Experience of therapist and sex of therapist . . Empathy Across Time—-Sex of Client and Sex of Therapist . . . . . . . Sex of Client (Hypotheses 5 and 5b) . . Issues . . . . . . . . . . . . Training Implications . . . . . . General considerations . . . . . Practicum students and first-year interns. . . . . . . . . . Second-year interns . . . . . . Senior staff. . . . . . . . . Becoming a therapist-—Males versus females. . . . . . . . . . In summary . . . . . . . . . Same—Sex Versus Cross—Sex Dyads . . . V 0 SUMMARY 0 O O . O O O O O O O O APPENDICES APPENDIX A. INFORMATION ON SELECTED MMPI CLINICAL SCALES B. CHARACTERISTICS OF D—PT CLIENTS . . . . C. RULES FOR RATING ACCURATE EMPATHY IN SPECIAL CASES 0 O O C C O O O O O O 9 D. RATER RELIABILITIES . . . . . . . . E. MMPI SCALED SCORES FOR CLIENTS OF BOTH GROUPS F. CLIENT CODE NAMES AND THE SESSIONS FROM WHICH THE TWO SEGMENTS WERE TAKEN . . . G. FOR EACH CLIENT: SEX OF CLIENT, EXPERIENCE LEVEL OF HIS OR HER THERAPIST, SEX OF THER- APIST, AS WELL AS MEAN, PEAK, AND BASAL EMPATHY OVER EARLY, LATE, AND SESSIONS COMBINED . . . . . . . . . . . Vii Page 72 72 73 73 73 73 74 75 76 77 79 82 85 87 92 94 96 97 98 APPENDIX Page H. AUDIBILITY O O O O O 0 0 O O O O O 100 I. CLIENT CODE NAME, AGE, VERBAL SCORE ON COL- LEGE QUALIFYING TEST (COT), TOTAL SCORE ON CQT. o o o o o c o o o o o o 10]. J. GENERAL CHARACTERISTICS OF THE DATA, AND ADDITIONAL STATISTICALLY SIGNIFICANT CORRELATIONS O O O O C O O O O O O 103 K. MEAN, PEAK, AND BASAL EMPATHY OF PRACTICUM STUDENTS, FIRST—YEAR INTERNS, SECOND-YEAR INTERNS, AND SENIOR STAFF EACH CONSIDERED SEPARATELY (ANALYSIS III) . . . . . . 104 L. EMPATHIC UNDERSTANDING IN INTERPERSONAL PROCESSES . O O O O O O O O O O O 110 REFERENCES 0 O O O O O O O O O O O O 0 ll 4 viii LIST OF TABLES Table Page 1. MMPI Means and Standard Deviations for Groups 1 and 2 (T Scores) . . . . . . . 21 2. E Test Results Between Group 1 and 2 Means on MMPI Scales. . . . . . . . . . . . 22 3. Mean, Peak, and Basal Accurate Empathy for Clients High and Low in Subjective Discom- fort Over Early, Late, and Early-Late Sessions Combined. . . . . . . . . . 41 4. Mean, Peak, and Basal AE for Experienced and Inexperienced Therapists Over Early, Late, and Early-Late Sessions Combined (Analysis I) . . . . . . . . . . . 49 5. Mean, Peak, and Basal AE for Senior Staff and Inexperienced Therapists Over Early, Late, and Early—Late Sessions Combined (Analysis II) . . . . . . . . . . . 49 6. Analysis of Variance of Peak AB in the Late Session by Sex of Client, Sex of Therapist, Experience of Therapist, and Client Group Status (Analysis I) . . . . . . . . . 51 E—l. MMPI Scaled Scores for Clients of Both Groups . 96 G-l. Sex of Client, Experience Level of His or Her Therapist, Sex of Therapist, as well as Mean, Peak, and Basal Empathy Over Early, Late, and Sessions Combined . . . . . . . . 98 I-l. Client Code Name, Age, Verbal Score on College Qualifying Test (COT), Total Score on CQT. . lOl K-l. Analysis of Variance of Mean Accurate Empathy in the Late Session by Sex of Client, Sex of Therapist, Experience of Therapist, and Client Group Status . . . . . . . . . 104 ix LIST OF FIGURES Figure 1. Mean Accurate Empathy in the late session broken down by therapist experience and client group status (Analysis I) . . . . 2. Basal Accurate Empathy in the late session broken down by therapist experience and client group status (Analysis I) . . . 3. Peak Accurate Empathy in the late session broken down by sex of client and client group status (Analysis II) . . . . . 4. Mean AB in the early session broken down by sex of therapist, client group status, and therapist experience . . . . . . . . 5. Basal AB in both sessions combined broken down by sex of client and therapist experience . . . . . . . . . . . 6. Peak AB in the late session broken down by therapist experience and sex of therapist . 7. Peak AB in late session as compared to early session broken down by sex of client and client group status (Analysis I) . . . . 8. Peak AE across time broken down by sex of client and therapist experience (Analysis II) . . . . . . . . . 9. Peak AE across time broken down by sex of therapist and therapist experience . . . lO. Basal AE across time broken down by sex of therapist. . . . . . . . . . . . 11. Peak AE in the late session broken down by sex of client and sex of therapist . . . Page 43 44 45 47 50 51 55 57 58 59 6O Figure Page K-l. Mean late session empathy broken down by therapist experience and client group status 0 O O O O O O O O O O O O 105 K-2. Basal late session empathy broken down by therapist experience and client group status . . . . . . . . . . . . . 106 K-3. Late session mean empathy broken down by therapist experience . . . . . . . . 106 K—4. Combined session basal empathy broken down by therapist experience and sex of client . . 107 K-5. Peak empathy across time broken down by sex of therapist and therapist experience. . . 108 K-6. Mean empathy across time broken down by therapist experience and client group status . . . . . . . . . . . . . 109 Xi CHAPTER I INTRODUCTION AND REVIEW OF LITERATURE Introduction Much research has been generated regarding facili— tative conditions offered by the therapist in the psycho- therapeutic relationship. Of the several factors, empathy has been called the most critical. It has been assumed that empathic communication by the therapist is an aid to client self-exploration and therapeutic depth. This study compares two somewhat similar groups of clients seeking psychotherapy. The first group manifests a clinically significant pattern of self—described symptoms of depression and anxiety (high subjective discomfort) while the second group represents a less pronounced pattern of such discomfort. Both groups will be compared with regard to the extent to which they elicit empathy from their therapists. This investigation raises such questions as: does degree of client subjective discomfort affect empathy -elicitation? Is there an interaction between therapist experience and amounts of empathy Offered to clients with varying degrees of discomfort? Do female clients elicit more empathy from therapists across experience levels? Do experienced therapists provide higher empathy than do inexperienced therapists? Is empathy offered higher in later therapy sessions than in earlier sessions? This research offers greater control and speci— ficity of client characteristics than is often available in psychotherapy analyses. Within the limits of available data resources, "matching" was utilized on selected important variables. Empathy Overview Empathy, the communicated understanding of another's experience, has generated much discussion and research, especially within the past decade. Rogers (1957) suggested it as one of three conditions "necessary and sufficient" for successful psychotherapeutic change. Carkhuff has been responsible for many studies in the area of therapeutic conditions, as well as two integra— tive volumes which include how facilitative conditions can improve human relations (Carkhuff, 1969). Truax and Wargo (1966) found that the higher the facilitative conditions of empathy, nonpossessive warmth, and genuineness, the greater the positive change in the client. They also discovered that both the average level of rated empathy, and the highest level of rated empathy were important to client improvement (as measured by psychological test data, diagnostic evaluations of patient change, and time spent out of the hospital Since initiation of psychotherapy). Truax and Carkhuff (1967) concluded that empathy was the central ingredient of the psychotherapeutic pro— cess. Rogers (1967) stated that Accurate Empathy (AE) and patient—perceived congruence were positively related to various indices of favorable outcome, and that "Accurate Empathy ratings would perhaps provide the most adequate and meaningful assessments of the general level of con— ditions in the relationship." Therapists who were high in facilitating conditions increased client self—exploration following confrontation (Anderson, 1968) and elicited high levels of patient experiencing and problem expression (Vander Veen, 1965). When compared to low functioning therapists, therapists high in facilitative conditions were also less susceptible to client manipulations (Carkhuff & Alexik, 1967). The successful client group had therapists who provided higher conditions (of empathy, regard, genuineness, and con- gruence) than the nonsuccessful group both in early and late periods of therapy (Schauble & Pierce, 1974). The authors concluded that the therapist has a "substantial influence" on the client's functioning level. Vesprani (1968) reported that college companions with high "D" and "Pt" MMPI scales were likely to be low in Accurate Empathy. Mullen (1969) found that experienced therapists were more consistent in offering empathy, and better at avoiding low empathy conditions than were inexperienced therapists. Safran (1973) reported several differences on the Edwards Personal Preference Scale (EPPS) between high and low empathic counselors, and concluded that empathic counselors were more people-oriented and were more positive about themselves. Caudillo (1972) found additional differences between high and low empathizers, as well as sex differences in this regard. Jones (1974), using the Carkhuff Scales, discovered that empathy related positively to an Index of Communication (£é.49) and toler- ance of ambiguity (£é.45), but related negatively to social introversion (re-.54) and order (£=-.55). Empathy-~Stable Characteristic or Situationally Defined? There have been conflicting opinions as to whether empathy is a stable characteristic of the therapist, or merely a behavior that varies with situations.1 A parallel issue is the extent to which the therapist determines his empathy levels. 1A complicating issue has been the fact that ver- bally empathic communication by the therapists were often in response to nonverbal behavior of clients. Shapiro (1968b) however found that empathy ratings of audio tapes were highly correlated (r=.70) with empathy ratings based on audio—visual recordings. He conceded that empathy was largely a verbally oriented scale, and confirmed the use- fulness of audio tapes as a "reasonable abstraction of the whole interaction." There is evidence that the therapist himself (and not the client) accounts for the level of AE (Truax, Wargo, Frank, Imber, Battle, Hoehn—Saric, Nash, & Stone, 1966; Truax, 1963), and that therapists were "remarkably con- sistent" over time (Schauble & Pierce, 1974; Melloh, 1965). Although Rogers originally conceived of therapeutic con— ditions as stable attitudes of the therapist, he (Rogers, 1967) discovered that empathy levels did not tend to stabilize until the 7th session (presumably until the therapist got a better "feel" for what the client's problems and needs were). Heck and Davis' (1973) results suggested that more complex counselors (those with higher conceptual function— ing as measured by incomplete sentences) expressed more empathy than less complex counselors, and that the level of empathy manifested was conditioned by a significant interaction effect between the type (high complexity versus low complexity) of counselor and client. They concluded that empathy was probably not a characteristic which remains constant across all stimulus conditions. Vander Veen (1965) too concluded that therapist behavior was a function of the therapist ang_the patient. Beutler, Johnson, Neville, and Workman (1973), as well as Gurman (1973), suggested that a therapist was neither consistent in empathy levels across sessions, nor within a single interview. Gurman also reported that the peak of high conditions within a particular hour seemed to occur in mid—late and late segments. This seemed to add further support to results of Karl and Abeles (1969) which indicated that certain variables appeared more frequently during certain segments of the hour. Accumulating research seems to point to the therapist as the primary determiner of empathy levels, with the patient a contributing factor. Gurman's (1973) results suggested a partial resolution with regard to the issue of empathy stability. Apparently, therapists are stable within a finite range, but vary within that range from moment to moment. Perhaps, "the more initially expressive the patient, the richer will be the material in which the therapist can anchor his empathic efforts" (Rogers, 1967, p. 308). Empathye—Criticism and Counter— Arguments Mixed results. Studies which offered mixed results with regard to therapeutic conditions (especially empathy) were the following: Mitchell (1971) found that an Immediate Relationship Scale2 was positively related to 2This scale was designed to assess the extent to which the therapist responds or fails to respond to direct and indirect references by the client to the therapist. It is a 6—stage scale rated by others, and Mitchell found it to be a reliable instrument capable of making signifi- cant and meaningful discriminations among therapists. conditions of high functioning therapists, but not related to positive outcome; Hogan (1969) developed a new empathy scale which shared much with common conceptions of "under- standing," but also correlated highly with "good" and "healthy"; Kurtz (1970) discovered that there was a sig— nificant correlation (£;.47) between client self—exploration and counselor empathy, but that six measures of empathy were unrelated to each other; Kurtz also found the Barrett— Leonard Relationship Inventory to be the best predictor of outcome (when filled out by the client after the 3rd session); Zimmerman (1973), however, failed to find a relationship between counselor skill in accurate perception of affect and client's perception of counselor empathy; Mullen and Abeles (1971) concluded that although high empathy over any stage is significantly related to posi- tive outcome, high liking and high empathy together did not predict outcome. Venema (1970) suggested that empathy with lower— class clients may be negatively correlated with length of stay in psychotherapy. It may be then that the rule "the more the better" does not apply to empathy in all situ- ations. Perhaps, it is only effective with certain clients at certain times during therapy. Empathy may "scare off" lower-class clients or psychotic clients who are not used to empathy or ready for it. Hammer (1968) noted that the use of interpretation may need to be varied depending on the diagnostic group of the client. Why not empathy as well? Validity. Many have raised questions concerning the validity of empathy as a construct (Langer, 1972; Rappaport & Chinsky, 1972; Chinsky & Rappaport, 1970; Kurtz & Grummon, 1972). Chinsky and Rappaport (1970) claimed that reliability estimates of the Truax AE Scale were related to the number of therapists being rated, and that raters were responding to some quality other than that which the scale defines. Their argument focused on Truax's report (1966) that there was no difference between AE ratings on nonedited tapes versus tapes with patients' statements edited out (5 between the two = .68). Furthermore, they maintained that any scale which was defined in terms of response to client statements cannot be measuring what it claims if equally good ratings were obtained without those client statements.3 Recently, Rappaport and Chinsky (1972) stated that empathy seems to be more highly correlated with other 3The way the scale is defined, client expressions are certainly needed to rate AE. However, there are probably some statements which are usually rated as reasonably empathic i.e. statements offering accurate reflection or the minimally facilitative conditions as defined by Carkhuff (level 3). Levels one and two do not represent mild degrees of empathy, but represent communi— cations which detract from the client's content and affect. dimensions (e.g. good-bad on a semantic differential, the number of words Spoken by the therapist) than with various measures all contending to measure the empathy dimension. The Accurate Empathy Scale correlated .68 with therapist "intensity" and "intimacy" (Truax & Carkhuff, 1967), and also correlated highly with Warmth and Genuineness (Shapiro, 1968a). Thus, it is suggested that empathy is not a spe- cific, unitary concept, and that ratings of it seem to assess some global attributes of the therapeutic inter— action. Though Chinsky and Rappaport's suggestions are useful, and their arguments provocative, there is a whole body of research which still seems to support empathy as a useful concept. Martin and Toomey (1973) offered a recent cross-validation of empathy as they discovered that empathic Ss tended to be field independent (as measured on the Embedded Figures Test). They suggested that empathy was related to psychological differentiation. Watts (1973) reported a high correlation between Carkhuff Discrimination Scores and measures of the 88' accuracy of perception of patients, and offered the Carkhuff Scale for use for rater selection. Bozarth and Krauft (1973) addressed some of the above validity and reliability questions. Specifically, they focused on the extent to which empathy ratings were in response to general therapist characteristics, and the lO extent to which high rater reliability was inflated by having more than one segment per therapist or by having few therapists in the study. with these issues in mind, Bozarth and Krauft had their raters judge "Good Therapist" and "Likability" following Accurate Empathy Ratings, and found significant but not high correlations among the three. They concluded that though some relation exists between empathy and more general counselor characteristics, empathy is relatively independent considering that the ratings were made at the same time. In addition, they found that judge reliability remained high (greater than .70) even with many therapists included in the ratings. That ratings were not inflated due to inclusion of more than one segment per therapist offered further contrary evidence to earlier criticisms (Chinsky & Rappaport, 1970; Rappaport & Chinsky, 1972). Perhaps the critics can be partially assuaged by those who have attempted an overview and integration of much research in the area of therapeutic conditions con- ducive to successful outcome. Matarazzo (1971), after mentioning several studies, concluded that . . . research is cohesive and nearly unanimous in suggesting that the conditions of warmth, accurate empathy, positive regard, and genuineness are impor- tant, although not the only, variables in determining depth of patient exploration and therapy outcome. They also appear to be important factors in deter— mining the effectiveness of the supervisor-student relationship. (Matarazzo, 1971, p. 900) ll Luborsky, Chandler, Auerbach, and Bachrach (1971) sup- ported "empathy" as a useful concept overall, and stated that "the therapist's empathy [and other related qualities] facilitates the patient's gain from psychotherapy." Rogers' (1967) collaborative work dealing with psychotherapy and schizophrenia suggested that if empathy was not a predictor of outcome by itself, it was a contributor to a favorable climate which made positive change possible. Patterson (1969) summarized by saying that empathic understanding (as well as nonpossessive warmth and genuineness) were well-established both theoretically and experimentally as being related to client self-exploration and various outcome criteria. In a recent article, Rogers (1975) indicated that "a high degree of empathy in a relationship is possibly the_most potent and certainly one of the most potent factors in bringing about change and learning." In summary, though having its critics, empathy has been widely accepted as being an important variable in any therapeutic encounter. To researchers who demonstrated a lack of relatedness between empathy and outcome, I say that empathy may be just one prerequisite to positive change. We have yet to uncover all of the other signifi— cant components. 12 Therapist-Client Interaction General Issues The consensus now seems to be that therapist and client are both influencing each other's behavior in therapy, with the therapist having a greater effect. Many have called for increased exploration of the ther- apist as he effects the client's potential for positive change. Pierce, Carkhuff, and Berenson (1967) suggested that clients will adjust to the facilitative condition level provided by their therapists. Persons, Persons, and Newmark (1974) reported that all clients showed improvement with psychotherapy, but clients were more responsive to counselors of the same sex. Hill (1975) agreed that same—sex pairs have more discussion of fee1~ ing. She also found an interesting interaction. Inexper— ienced males and experienced females were more active and empathic than experienced males and inexperienced females. She concluded that counselors responded differently to different clients (depending on whether the clients were male or female). Rosenzweig and Folman (1974) stated that the therapist was the most reliable judge of who he can best help, and suggested an examination of therapist attitudes as they relate to different types of patients. Love (1971), on the other hand, stressed that the client had a greater influence on the therapist than is generally realized. Lauver, Kelly, and Froehle (1971) l3 concur. They examined how the therapist's verbal behavior was modified by varying "reaction time latencies" of the client. They discovered that clients can affect changes in speech and silence behavior of counselors. While similarity of value and cultural systems between client and therapist seems to aid formation of the therapeutic relationship and thus be a plus factor when predicting outcome (Snyder, 1961), dissimilarity was more frequently associated with counseling success in terms of needs or styles (Bare, 1967). Counselors had more success when they were unlike their clients on "original thinking," "responsibility," and "vigor." Snett (1972) discovered that distinguishing between conflict and defensive similarity was important. Conflict similarity was found to be negatively related to liking. The well-known study by Betz (1967) found that certain therapists (who could be differentiated via an interest inventory) did better with schiZOphrenics than other therapists, and generated much research in the area of therapist—patient interaction. Apparently, physicians who were judged to be more effective with schizophrenic patients (dubbed "A" type) had more problem solving interests than mechanical interests. Although there has been cross—validation of the AeB dichotomy, Stein, Green, and Stone (1972) found no support for the pairing interaction as it affected outcome. 14 In fact, they reported that both_"A" and "B" type phy— sician therapists held more favorable attitudes toward neurotic and middle-class patients. Beutler, Johnson, Neville, and Workman (1972), although supporting Betz's original findings, found no relation between empathy and judged improvement or length of hospitalization. Scott and Kemp (1971) reported that "B" type therapists did elicit greater client depth of exploration, but found no significant relationships between therapist A—B scale scores and their empathy, warmth, and genuine- ness offered to neurotic outpatients. Bergin and Suinn (1975) stated that the A-B dichotomy was too simple a concept to deal with the complexity of the therapist— client interaction. They suggested a closer inspection of studies in the area. Different patient types apparently have varying needs, and therefore value different things in therapy. Betz (1967) suggested that safety and not communication was the primary goal of schizophrenics. Thus schizo— phrenics will likely prosper with a therapist who is able to encourage an atmosphere of security. The neurotic, although certainly needing a trusting relation— ship, may place more importance on being understood. Fernbach (1973) offered another area where therapist-patient interactive components need to be considered. He suggested that authoritarian clients 15 prefer directive therapists. Blumberg (1972), in a similar approach, reported that highly dogmatic Ss tended toward positive change with "leading" interviewers but negative change with "following" interviewers. Lerman (1963) established that therapists judged less competent were more conflicted and anxious. Further— more, she found partial confirmation that the therapist's dependency anxiety was related to both client and therapist. Bergin and Solomon (1970) reported a moderate inverse relationship between therapist's empathy and therapist's MMPI disturbance level. More specifically, "D" and "Pt" scales correlated negatively with empathy. This result was replicated in a further study (Bergin & Jaspar, 1969). On the other hand, Jones (1974) did not obtain statistically significant results in this regard though his findings were in the expected direction. Scher (1975) found that clients seeing experienced counselors reported better outcomes (with facilitative conditions held constant). His other results indicated that neither sex nor activity of therapeutic participants contributed to outcome. Meltzoff and Kornreich (1970), in an authoritative review of psychotherapy research, concluded that no clear relationship existed between sex of patient and outcome, but that counselor experience did make a difference. 16 Thus, it is clear that a certain group of thera- pists seem to do exceptionally well with a great per- centage of their clients. Furthermore, there is a need to compare client types (based on needs, diagnostic groups, personality characteristics, etc.) with per- sonality characteristics, styles, and need systems of therapists. Client Discomfort and/or Disturbance, As Related to Psychotherapy Process and Outcome Strupp, Wallach, Jenkins, and Wogan (cited in Bordin, 1974) found the client's initial disturbance to be related to the therapist's ratings of improvement (5?.63). Snett (1972) mentioned that therapy prognosis ratings were positively related to the degree to which patients were perceived as being motivated for therapy. Mullen (1969) discovered that the "successful" outcome group (defined as changed from higher MMPI scores to lower MMPI scores following therapy) had significantly higher MMPI scores preceding therapy than the "unsuccess- ful" group (who demonstrated no change or increased scores on the MMPI following therapy). Thus it appeared that the successful group was more disturbed prior to therapy as compared to the unsuccessful group who had MMPI scores more within a "normal" range. The two groups were more alike_following therapy than before, suggesting that success reflected getting the more disturbed client to l7 improve. Psychotherapy was less effective with the group with initially low MMPI scores. Meltzoff and Kornreich (1970) cited several sources which offered evidence that initial discomfort, initial symptoms, or initial maladjustment scores were unrelated to outcome (Gliedman, Stone, Frank, Nash, & Imber, 1957; Stone, Frank, Nash, & Imber, 1961; Page, 1953). On the other hand, Ewing (1964) found the largest change in those with the greatest dissatisfaction with themselves. He also found tentative support for the hypothesis that changes in a problem area were greatest for clients whose initial status in that area deviated most from normal persons. Campbell and Rosenbaum (1967) found that outpatients with high initial distress reported more relief over four weeks of therapy, and Levis and Carrera (1967) suggested that Ss with the highest MMPI scores improved the most. Truax and Wittmer (1971) discovered that the degree of improvement was unrelated to the initial level of adjustment, but was significantly related to change in discomfort scale (£?.67). They maintained that the greater the focus on the anxiety source, the greater the improvement. Therapists may like anxious—depressed clients because they (the therapists) can see maximum change with minimum effort, i.e. get the discomfort to change fairly quickly. 18 Frank, Gliedman, Imber, Nash, and Stone (1957) stated that . . . the worse a patient says he feels, the more likely he is to accept treatment. The degree of the patient's distress is an indicator both of the strength of his motivation for treatment and of his willingness to communicate with the therapist. (p. 293) Furthermore, they cited Dollard and Miller who maintained that the more miserable the patient, the greater his moti- vation to learn effective responses which will reduce his distress. Gliedman, Nash, Imber, Stone, and Frank (1958) used placebos in conjunction with short-term psychotherapy in an attempt to reduce symptoms. They defended the utility of placebos, and suggested that symptoms of anxiety and depression were most susceptible to placebo effects. CHAPTER II EXPERIMENTAL DES IGN Subjects Both therapist and client data, as well as tapes of therapeutic interactions, are on file at the Michigan State University Counseling Center. An attempt was made to select two comparable popu- lations of clients who sought therapy. One group of clients was chosen to represent a specific symptom con- stellation consisting primarily of depression and anxiety. The other group chosen was considerably lower (to a sta- tistically significant degree) on depression and anxiety, but reported essentially similar behavior in other respects. Those subjects who had T scores greater or equal to 65 on both scale 2 (depression) and scale 7 (psychasthenia) of the MMPI were included in group 1 (high subjective dis— comfort group). No cases were used if any validity scales (L, F, K) were greater or equal to 70. Using these cri— teria, 20 clients were selected. These 20 clients (12 females and 8 males) were seen by 13 male therapists and 7 female therapists. The 19 20 experience level of the therapists was as follows: 7 staff, 7 second—year interns, 4 first—year interns, and 2 practicum students. A comparison group (group 2) was selected which closely approximated group 1 clients in terms of sex of client, sex of therapist, and experience level of thera- pist. Again, cases were not accepted which had any validity scales greater or equal to 70 (there was one exception). These individuals (also 12 females and 8 males) were seen by 15 male therapists and 5 female therapists who in turn consisted of 6 staff, 5 second— year interns, 7 first-year interns, and 2 practicum stu— dents. This group was the best "fit" in the available client pool. Table 1 contains the means and standard deviations of the MMPI clinical and validity scales for both client groups. E tests were performed to ascertain the extent to which the groups differed or were alike. Table 2 (p. 22) summarizes the E test values. As one might expect based on the selection pro— cedure, group 1 was significantly higher (p < .01) than group 2 on "D" and "Pt." An additional finding was that group 1 was also higher than group 2 (B < .05) on "Pa" and "Si." The meanings of these additional differences will be discussed later. There were no significant 21 m.HH mm.m wa.oa mm.n mw.m v.NH v.0H mm.m mm.m mm.h me.h m.m oH.m .Q.m mm.mm m.mm mm.mm mm.em m.mm o.am >.Ho o.mm H.vm m.Hm >.mm m.mm H.ev coo: N mzouw m.HH m.mH Hm.m om.m NH.m m.vH mm.m mo.m ve.o mv.n no.o Hw.v vv.w .o.m m.Hm m.mm m.mm m.H> N.Hm o.mm H.vm m.Ho m.vn m.mm e.om H.nm v.mv coo: mamom o m w n m m w m m H M A q HAZE H msouo Amouoom Ev N can A mmsouo How mcofluwfl>on pumpccum Ugo memo: Has: A magma 22 differences either on the validity scales or the other clinical scales. One further 5 test was performed com— paring the number of sessions group 1 and group 2 clients were seen in therapy. The difference was not significant (3 was less than 1). Table 2 2 Test Results Between Group 1 and 2 Means on MMPI Scales L .97 nonsignificant (n.s.) F .79 n.s. K 1.36 n.s. Hs by inspection n.s. D 8.63 (P. < .01) Hy .89 n.s. Pd .73 n.s. Mf by inspection n.s. Pa 2.55 (B < .05) Pt 5.39 (E < .01) Sc 1.46 n.s. Ma .83 n.s. Si 2.45 (g < .05) Because of the elevations on scales 2 and 7, one might make a case that those in group 1 were clinically "sick" or more disturbed. On the other hand, Taulbee (1958) stated that "painful affects which the neurotic experiences are signs of strength within the personality." Furthermore, the groups are not different on any of the validity scales. The lack of difference between the mean 23 T scores on L is not surprising. It appears that there was no overt attempt to create a false impression by either group. The selection procedure (by using F scale scores less than or equal to 70) deliberately omitted 55 who were either reporting rather bizarre sensations, strange experiences or who were "faking bad." Zuckerman and Monashkin (1957) found self-acceptance positively cor- related with MMPI "K" scale scores2 and negatively cor— related with MMPI "F" scale scores.3 In addition, Canter (1960) reported that the value of K dropped as the degree of disturbance increased. Again, that the two groups did not differ on these scales supports the contention that the groups do not vary appreciably in terms of maladjustment. The reported subjective discom- fort seems to be the major difference between the groups. 1L scale items are not frequently answered in the critical direction by college students, but the content "refers to denial of aggression, bad thoughts, weakness of character or resolve, poor self—control, prejudices, and even minor dishonesties" (Dahlstrom, Welsh, & Dahl- strom, 1972, p. 109). 2The K scale is often used as an indicator of test-taking attitude, and suggests the state of personal defensiveness. 3The F scale was "designed to detect unusual responding or atypical ways of answering the test items" (Dahlstrom et al., 1972, p. 113). 24 In addition to D and Pt, group 1 was significantly higher than group 2 on Pa. Theoretically speaking, this does not pose any major difficulty because Pa fits psy- chodynamically with D-Pt. The clusters of Pa are per- secutory ideas, naivete (moral virtue), and poignancy (intensity of feeling). At sub-clinical levels (group 1's mean = 61.2), sensitivity and worry are more probable than paranoia. Zuckerman and Monashkin (1957) reported that patients high on the Pa scale may be just admitting negative traits in themselves. Hartley and Allen (1962), in a factor analytic study, found that the Pa, Pt, and Hy scales related positively to "anxious oversensitivity." Group 1 was also significantly higher on Si (social introversion). This is a relatively new scale, and often not used as a clinical scale but as a measure of the other scales' effect on social withdrawal. Thus, the scale is more a reaction to distress than distress itself. The scale's items include those with strong self— depreciatory valences. The MMPI Handbook (Dahlstrom et al., 1972) described men with high "0" (Si) as being modest, while women are called "shy, self—effacing, and sensitive." The group 1 (N=20) Welsh code is 27'84936f915/ FK/L:. Males in this group may be described as depressed, somewhat anxious, neurotic, somewhat obsessional, more likely to show feelings, and somewhat better bets for 25 therapy (because needy and "hurting") than group 2 members. In addition to the above, females in group 1 may be described as conscientious with traditional values. The overall profile of group 1 has clinical peaks (greater than 2 standard deviations beyond the standardized mean) on scales D and Pt, and several other moderate elevations (greater than 1 standard deviation beyond the standardized mean). It is a high normal profile with disturbances taking the form of depressive and anxiety manifestations, especially representative of low self—esteem. The Welsh code for group 2 (N=20) is 93721 762015/ FK/L:. Clients in this group may be described as alienated, and having home conflicts and interpersonal problems. This group is overtly functioning better than group 1, but there is more denial and emptiness. They verbalize feelings rather than demonstrate them. The group 2 profile is a high normal profile with two moderate elevations (greater than one S.D. above the mean). Zuckerman and Monashkin (1957) raised the possibility that extremely self—accepting 88 may be maladjusted, but defensive. In any case, Zucker and Manosevitz (1966), citing Dahlstrom and Welsh (1960), mentioned that "workers with the MMPI have emphasized that the diagnostic utility of the instrument extends below the T = 70 level." For the most part, groups 1 and 2 are relatively similar, but demonstrate differing manifestations of distress. 1 is -_—_..___—_-_._._..._—- .1: ._ .— 26 Method Selected three-minute segments of two therapy sessions of each client will be rated on Accurate Empathy using Carkhuff's Empathic Understanding in Interpersonal Processes: A Scale for Measurement (Carkhuff, 1969). Karl and Abeles (1969) suggested that therapist approach was greatest during the second and third ten-minute inter- vals of each session, while Gurman (1973) specifically found offered empathy to be highest during mid—late and late segments. With both of the above studies in mind, segments of the 35th to the 38th minute of each therapy session will be examined. The two sessions selected for sampling will be the 1/4 point and the 3/4 point ofeach completed therapy. It is hoped that this will optimize the potential for finding high empathy levels, as well as facilitate examination of empathy as it varies over time. The judges' ratings will be averaged, as will the scores from the two tapes. Hypotheses Hypothesis 1: Clients high in pre-therapy subjective discomfort (those who had high depression and psychasthenia MMPI scores) will elicit significantly higher Accurate Empathy across all therapist experience levels than will clients low in subjective discomfort (those with significantly lower levels of depression and psychasthenia). 27 There is some evidence that the healthier the patient is initially, the better the outcome (Luborsky, Chandler, Auerbach, & Bachrach, 1971; Barron, 1956). On the other hand, others found more improvement in those clients who seemed initially "sicker" than com— parison clients (Mullen, 1969; Strupp et al., 1963). Luborsky et a1. (1971) also discussed several patient factors which were most often significantly associated with improvement, and among these factors were anxiety and motivation. Plyler (1965) stated that "D-Pt" profiles in particular may indicate readiness for counseling. Therapists usually prefer well—motivated clients since they are perceived to improve faster (Meltzoff & Kornreich, 1970; Snett, 1972). Indeed, Gallagher (1956) found that discomfort scales on the MMPI showed the greatest tendency toward change. More specifically, F, K, Hs, D, Pt, and Si scales changed due to therapy, with greater changes occurring in the last three scales. Bergin concurs. He asserted that: Certain MMPI scales repeatedly yield evidence that they are able to detect client change. Among those scales that appear to provide consistent validity as change indices are D, Pt, and Sc. (Bergin, 1971, p. 260) Because of initial depression and anxiety, group 1 clients will appear more self-dissatisfied and more motivated for change than will group 2 clients. 28 Therapists may sense the "pull" for succorance and empathy, and "give" more. Stone, Frank, Nash, and Imber (1961) found that anxiety and depression subscales showed rapid initial improvement and were the only ones to Show significant change at six months. The therapist then may become especially motivated himself when he becomes aware of the client's overt distress, and of the possibility of reducing that discomfort in a reasonably short span of time. Hypothesis 2: Experienced therapists will offer significantly higher Accurate Empathy than inexperienced thera— pists (a) for both experimental and comparison groups, and (b) over early and late psychotherapy sessions. Hypothesis 2 (restated): Senior staff and second-year interns (N=25) will offer significantly higher Accurate Empathy than practicum students and first—year interns (N=15) (a) across both clients high in subjective discom- fort (group 1) and those low in such discomfort (group 2), and (b) will occur over both early and late psychotherapy sessions. Experienced therapists do seem to obtain more favorable client outcomes overall than do the inexper— ienced therapists (Meltzoff & Kornreich, 1970; Scher, 1975; Bergin, 1971). In addition, experienced therapists offer a higher degree of empathy than do the inexperienced therapists (Rogers, 1975; Mullen & Abeles, 1972). 29 Fuller (1961) however found no difference in amount of feeling expressed in the presence of exper- ienced and inexperienced counselors. Furthermore, Hill (1975) suggested that experienced male therapists do not provide high levels of Accurate Empathy. Thus, Hypothesis 2 will hopefully clarify these puzzling find- ings, and ultimately reconfirm earlier results. Exploratory Hypothesis 3 If both Hypotheses l and 2 are confirmed, then additional questions will be considered. The following four statements are to be considered the plausible possi— bilities that could occur regarding therapist experience and client discomfort. The exact nature of the relation— ship will not be predicted, and these should be considered exploratory (and potentially incompatible). Statement 3a: There is no interaction between client discomfort and therapist experience with regard to AE. experienced therapists rated levels :::::::::: inex erienCed of AB 7 p PA* A therapists group 2 group 1 The above statement follows from a strict inter— pretation of Hypotheses l and 2. That is, although both groups of therapists offer greater AE to group 1 than to 30 group 2 clients, the experienced therapists will still offer higher AE than will inexperienced therapists over both client groups. Statement 3b: Experienced therapists provide higher AE to group 1 clients than to group 2 clients. F”’//,,g experienced therapists rated levels - i e rie d of AE n xpe nce —¢ : therapists group 2 group 1 This statement suggests that experienced thera— pists will provide higher levels of AE to group 1 clients than they will to group 2 clients, whereas inexperienced therapists will provide similar levels of AE to both client groups. Statement 3c: Inexperienced therapists provide lower AE to group 1 'clients than to group 2 clients. e I experienced therapists rated levels ‘\“\\1\\~ - . 1nexper1enced Of AE therapists group 2 group 1 This statement suggests that inexperienced thera— pists provide lower levels of AE to group 1 clients than to group 2 clients, while experienced therapists will provide similar levels of AE to both client groups. The inexperienced may become anxious themselves when working 31 with group 1 clients, and become less understanding. Experienced therapists, on the other hand, have been shown to offer consistent levels of AE when compared to inexperienced therapists (Mullen, 1969). Statement 3d: Experienced therapists will provide higher levels of AE to group 1 clients than to group 2 clients, whereas inexperienced therapists will provide higher AE to group 2 clients than to group 1 clients. */”/’/,12 experienced therapists rated levels “\“\~\\1‘ inex erienced of AE p : : therapists group 2 group 1 Group 2 members may be easier to work with. Less experienced or less empathic therapists will probably prefer these, whereas group 1 may elicit more empathy from experienced therapists who will not be turned aside by the anxiety of the client. Good therapists will be able to see the behavioral signs for what they are, and not be "pushed" into withdrawing from the client. Hypothesis 4: Accurate Empathy offered will be significantly higher in later therapy sessions than in earlier therapy sessions. Presumably, empathy will be more accurate and more frequent after the relationship has been more estab- lished. That is, the therapist will "know" the client 32 better, and be more likely to understand the client's experiences. This is supported by Rogers (1967) who suggested that AE would be more stable after the 7th session. Recently, however, Rogers (1975) concluded that "the degree of empathy which exists and will exist in the relationship can be determined very early, in the fifth or even the second interview." This seeming contra- diction will be explored. \ Hypothesis 5: Female clients will elicit higher AE across thera- pist experience levels than will male clients. In terms of traditional sex roles, it is more acceptable for females to express anxiety and depression, and to elicit succorance. Fuller (1961) found that female clients were judged to have expressed more feeling than male clients both in intake and the first interview. To the extent to which therapists respond to these cul- tural stereotypes, they will offer higher levels of AE to females than to males. Exploratory Hypothesis 5b If Hypothesis 5 is confirmed, the level of AE provided to female clients by inexperienced therapists will be explored. Previous research (Mullen, 1969) suggested that experienced therapists were more consistent in offering AE than were inexperienced therapists. This 33 difference may be due to a variety of factors. Here we will explore whether inexperienced therapists provide higher levels of AE to female than to male clients. Hypothesis 5b: Female clients will elicit higher levels of AE from inexperienced therapists than w111 male clients. Graphically, Hypothesis 5b suggests: - - experienced therapists rated levels «“\\\\~\~ ' ° of AE 1nexper1enced . 1+. : therapists female male clients clients Instruments MMPI Certainly it is not necessary to delineate all of the research that has been associated with the MMPI over the years. The reader is referred to Basic Readings of the MMPI in Psychology and Medicine (Welsh & Dahlstrom, 1956), as well as An MMPI Handbook (Dahlstrom, Welsh, & Dahlstrom, 1972) for comprehensive coverage. Suffice it to say that the instrument has had extraordinary heuristic power, and a great number of cross—validations. A few of the most relevant studies will be mentioned here. While much work has been addressed to diagnostic use of the MMPI profiles (Marks & Seeman, 1963; Gilber- stadt & Duker, 1965), Meikle and Gerritse (1970) found that only 16.8% of patients could be classified using 34 the Marks and Seeman cookbook. Furthermore, only 27.2% of patients could be diagnosed via Gilberstadt and Duker rules, and only 36.3% could be classified if the rules were combined with no overlap. Meikle and Gerritse con- cluded that the rules were not applicable for the majority of cases in a psychiatric unit. The plethora of sub—scales in the appendix of Dahlstrom et a1. (1972) gives ample evidence of uti1i~ zations of the MMPI to answer specific questions. The 11 "experimental" scales now included in most MMPI com- puter analyses offer additional evidence of relatively new developments. In 1960, Kleinmuntz sought to identify adjusted from maladjusted students via 43 items on the MMPI. Using a cutoff of 15, he was able to achieve a 95% "hit" rate. (Kleinmuntz, 1962, has conveniently organized an annotated bibliography of MMPI research among college students.) Boerger, Graham, and Lilly (1974), noting the frequent emphasis of two-point code types, argued that single scale types can be a meaningful interpretive approach to the MMPI. Cooke (1967) used a regression formula which reliably replicated (£=.9l) clinician's ratings on MMPI profiles with college males. Her overall hit rate was 76%, and higher for the nonpsychiatric group. 35 More recently, Mezzich, Damarin, and Erikson (1974) offered a simplified regression formula for dif— ferential diagnosis of depressive states from other psy- chiatric conditions. They cautioned that base rates for any group under study be attained before use of the formula can be reliable. A regression formula was devised by Plyler (1965) to discriminate groups of male college students who had been previously diagnosed as having vocational, edu- cational, or emotional problems. He not only examined the usual MMPI scales, but the experimental scales as well. He discovered that "D" was one of the three most discriminating scales. After citing a number of studies, Plyler suggested that the "MMPI may have considerable utility in studying personality differences between diagnosed groups of counseled college students" (p. 8). He examined high and second high point scales, but only when T scores were greater or equal to 55. There were many further studies which supported the validity of the MMPI. A frequently asked doubt con— cerning the MMPI, however, is the extent to which self— reports are reflected in actual behavior. Drake and Oetting (1959) found consistent patterns of behavior associated with certain profile two—point codes. Peers described males high on D and Pt as 36 "tense, indecisive, unhappy, worrying a great deal," and females as "anxious, depressed, lacking in self- confidence, socially shy and insecure." Black (1956a, 1956b), Goodstein (1956), Drake (1954, 1956), Guthrie (1956), and Mello and Guthrie (1958) found particular personality characteristics of people with different profile high points and configur- ations. Gynther, Miller, and Davis (1962) cited a study which, using a self-reported Interpersonal Check List (ICL), had 80% agreement between the self-ratings and pooled ratings of an individual's behavior by others (provided one deals with nonpsychiatric group). Self— ratings then (like the MMPI) are a useful assessment procedure. Further support is offered by Mello and Guthrie (1958). They concluded that "there are differences in behavior predictable from MMPI profiles." Heath and Korchin (1963) found considerable congruence between clinical and self trait and state evaluations. A Scale for Measurement of Empathic Understanding in Interpersonal Processes The above scale was discussed and outlined by Carkhuff (1969). Five levels were calibrated with level 5 representing the highest level of empathic responding. This scale was derived in part from "A Scale for the Measurement of Accurate Empathy" which has been 37 widely validated (Truax & Carkhuff, 1967; Carkhuff & Berenson, 1967). Various studies have used the Carkhuff Scale and found it preferable to the Truax Scale (Langer, 1972; Kurtz, 1970; Watts, 1973). Carkhuff (1969) dealt with theoretical, research, and training aspects of facilitative conditions. He indicated that Accurate Empathy consisted of a high discriminative factor as well as an activity component. High level communicators had high discrimination scores (Carkhuff, Kratochvil, & Frill, 1968), emphasized the experiential, and knew how to employ their discriminations. Carkhuff (1969) suggested that level 3 was the minimally facilitative level of empathy. At this level, the helper's comments were interchangeable with those of the helpee. When therapist responses added to the content and affect of the client, he had offered higher levels (levels 4 and 5). When he subtracted or failed to reach the same level as the client himself, the therapist was functioning at levels 2 or 1. Thus, no matter where the client is functioning, the therapist should at least be at that level (and hopefully higher) for the client to improve. Since the ratings of AE are related so critically to the client's level of functioning, raters must relate the client's response to therapist empathy. Patient-thera- pist-patient (PTP) statement groups will be examined to see 38 whether clients change the subject, increase self- exploration, become silent, or withdraw following the therapist's response to the client's initial statement. In addition to the above studies, Hill (1975), Jones (1974), and McNally and Drummond (1974) all used Carkhuff's scales profitably. It seems that the Scale for Empathic Understanding is a widely used tool for assessing Accurate Empathy. CHAPTER III RESULTS General Considerations Analyses of variance were performed using four independent variables: client's group status (grst), sex of client (sexcl), sex of therapist (sexth), and experience level of therapist (expth). The dependent variables were the following: (1) the mean1 Accurate Empathy score in the early session (MAEl), (2) the mean AE score in the late session of therapy (MAE2), (3) the mean AE score of both sessions combined (MAE3), (4) the peak2 or highest AE achieved in the early session (PAEl), (5) the peak AE achieved in the late session (PAEZ), (6) the peak AE achieved in both sessions combined (PAE3), (7) the basal or lowest3 AE offered in the early lMean is derived by averaging AE ratings of each rater for a particular tape segment, and then averaging over both raters. 2Peak is derived by taking the highest AE rating of each rater for a particular tape segment, and then averaging over both raters. 3Basal empathy is derived by taking the lowest AE rating of each rater for a particular tape segment, and then averaging over both raters. 39 4O session (LAEl), (8) the basal AE offered in the late session (LAE2), and finally, (9) the basal AE offered in both sessions combined. Where applicable, Pearson product—moment correlations were utilized as well. In addition to the described dichotomization of therapist experience (senior staff plus second-year interns versus first-year interns plus practicum stu- dents--Analysis I), further analyses were performed with two additional categorizations of therapist experience. Analysis II examined differences between senior staff versus the other three experience levels combined, and Analysis III examined the differences among all four experience levels considered separately (staff, second- year interns, first-year interns, practicum students). While the focus will be on Analysis I, discussion of Analysis II will be presented where appropriate. Analysis III provides rather complex results which are presented and discussed in the appendix, as well as in the section on "Training Implications." The reader should note that, for testing purposes, hypotheses occasionally had to be broken down into more specific components. 41 High Versus Low Client Subjective DIScomfort Hypothesis 1: Clients high in pre—therapy subjective discomfort (those who had high depression and psychasthenia MMPI scores) will elicit significantly higher Accurate Empathy across all therapist experience levels than will clients low in subjective discom- fort (those with significantly lower levels of depression and psychasthenia). Client Discomfort Main Effects Hypothesis 1a: Clients high in subjective discomfort will elicit significantly higher Accurate Empathy than will clients low in subjective discomfort. Table 3 Mean, Peak, and Basal Accurate Empathy for Clients High and Low in Subjective Discomfort Over Early, Late, and Early-Late Sessions Combineda Group 1 Group 2 MAEl 3.13 3.11 MAE2 3.22 3.14 MAE3 3.18 3.12 PAEl 3.65 3.69 PAE2 3.81 3.61 PAE3 3.73 3.65 LAEl 2.51 2.53 LAE2 2.64 2.60 LAE3 2.58 2.56 aNo differences between group 1 and group 2 means were statistically significant. 42 As summarized in Table 3, one may see that no statistically significant difference exists between group 1 and group 2 means for mean, peak, or basal empathy. This lack of difference is true for early, late, and early-late sessions combined. There is how— ever a statistically significant Pearson correlation (5 = -.29; p < .05) that suggests that group 1 status is related to greater peak empathy in the late ses- sion. But for this exception, Hypothesis la can be rejected. Client Subjective Discomfort and Therapist Experience Hypothesis lb: Clients high in subjective discomfort will elicit significantly higher Accurate Empathy across all therapist experience levels than will clients low in subjective diScomfort. To test this hypothesis, analysis of variance 2-way interactions (client discomfort by therapist experience) were examined. Eaply sessions provide no statistically significant results. However, experienced therapists seeing anxious-depressed clients have higher 132g session mean empathy than do the inexperienced therapists with similar clients. There is no difference in mean empathy offered to other clients (see Figure l). 43 3 35__ experienced therapists 3.3 “ mean AE 3.2 -- in late 3.1 ‘— session 3'0 __ inexperienced therapists group 1 group 2 Clients Figure 1. Mean Accurate Empathy in the late session broken down by therapist experience and client group status (Analysis I) (statisti— cally significant interaction using ANOVA; p < .05) Also in the late session, experienced therapists are better able than the inexperienced to avoid low basal empathy with anxious-depressed clients. Surprisingly, the inexperienced are better able than the experienced to avoid low basal empathy with clients low in subjective discomfort (see Figure 2). Sessions combined Show a similar but stronger pattern (p < .01). Thus, one may conclude that clients who are high in subjective discomfort receive significantly higher empathy from experienced therapists, but not from inexperienced therapists. Hypothesis lb is partially confirmed. 44 2'75‘I inexperienced therapists 2.7 4' Basal 9B 2 . 6 + In late 4 . . session 2'5 experienced therap1sts 2.4 ‘— 2.3 j group 1 group 2 Clients Figure 2. Basal Accurate Empathy in the late session broken down by therapist experience and client group status (Analysis I) (statistically sig- nificant interaction using ANOVA; p_< .05) Summary of Hypothesis 1 When examining the mean or peak empathy in the gagly session of therapy, there is little evidence to support the hypothesis that group 1 clients receive higher empathy than do group 2 clients. When examining the 1333 session of therapy, the results become more com- plex. There is no main effect between group 1 and group 2 clients on mean empathy. There is however a statisti- cally significant interaction of client group by exper- ience of therapist. There is no difference between experience levels on mean empathy offered to group 2 clients, but experienced therapists offer significantly higher mean empathy with group 1 clients than do the inexperienced therapists with such clients. 45 Experienced therapists are also better at avoid— ing low basal empathy with group 1 clients than are the inexperienced therapists. For both sessions combined, there is no difference between groups 1 and 2 for mean or peak empathy, though experienced therapists are still better than the inexperienced at avoiding low basal empathy with clients high in subjective discomfort. Client Subjective Discomfort and Sex of Client There exists a trend (p < .10) that suggests that female clients who are not anxious—depressed receive slightly lower late session peak empathy than females with high subjective discomfort, whereas male clients who are low in subjective discomfort receive dramatically lower late session peak empathy than males high in such discomfort (see Figure 3). Peak 3'8 2g 3.7 female clients the late 3'6 sess1on 3.5 male clients group 1 group 2 Clients Figure 3. Peak Accurate Empathy in the late session broken down by sex of client and client group status (Analysis II) (statistical trend; E < .10) 46 No statistically significant differences result either in mean or basal empathy regarding this inter- action. Client Subjective Discomfort and Sex of Therapist There are no statistically significant findings regarding the client subjective discomfort-sex of thera— pist interaction in mean, peak, or basal empathy, and this was true for early, late, and combined sessions. Analysis II provides an interesting 3—way interaction. In the EEEIX session, there is a strong trend that sug- gests that male senior staff and inexperienced female therapists offer higher mean empathy to group 1 clients, while female senior staff and inexperienced male thera— pists offer higher mean empathy to group 2 clients (see Figure 4, parts 1 and 2). Experience of Therapist Hypothesis 2: Experienced therapists will offer significantly higher Accurate Empathy than inexperienced thera- pists (a) for both experimental and comparison groups, and (b) over early and late psychotherapy sessions. Hypothesis 2 overlaps with the newly specified Hypothesis lb. The reader is referred to page 42 for the reporting of additional relevant findings. 47 I Female Therapists (Part 1) 3.4~— 3-3‘I . . Mean exper1enced therap1sts AB 3 2—— in ’ ear¥Y 3.1—- seSSIOn 3.0‘v 2.9‘h inexperienced therapists group 1 group 2 Clients II Male Therapists (Part 2) 3.4—) Mean 3.3-- AB in 3.2 early 3 l~~ inexperienced therapists session ° 3.0‘t experienced therapists 2.9-“— group 1 group 2 Clients Figure 4. Mean AB in the early session broken down by sex of therapist, client group status, and therapist experience (statistically signifi— cant trend using ANOVA; p’= .057). 48 Experience of Therapist Main Effects Tables 4 and 5 suggest that experienced thera— pists tend to offer higher late session ngH_empathy than do the inexperienced therapists, and that senior staff also tend to offer higher late session mggp_empathy than do the inexperienced therapists. Correlational data corroborates the above. Greater therapist experience is significantly related to higher late session peak empathy (£'= .33; p < .05), and tends to be associated with higher late session mean empathy as well (£.= .23; p < .10). Thus, in general, experienced therapists offer empathy levels as high or higher than do inexperienced therapists. Therapist Egperience and Sex of Client In early and late sessions combined, experienced therapists avoid low basal empathy with female clients, but not with male clients, while inexperienced therapists avoid low basal empathy with male clients but not with female clients (see Figure 5). No therapist experience by sex of client inter— actions are statistically significant for mean or peak empathy, either early or late. _ .wocmoHMHcmHm mo Hw>wa mo. gnu pm 1 mapomxm we memos cmm3uon mogoHDMMHQQ .oa. vrm “pcoup uCMOHMHgmHm >HHo0Humflpmum CH mpasmmu magma cmokumn monouommaoo .oa. v m “pcmnu #SDOHmagmHm SHHMOHumHumum CH mpHSmoH magma coo3¢m£ DDSDHDMMHQD mm.m Hm.m mmmq mm.m om.m mmfiq mm.m mn.m mmfiq mm.~ mm.m mmmq mm.m mv.m Hmdq om.m mm.m Hmdq mm.m m>.m mm¢m mm.m mh.m mmdm aw.m mm.m nmmam mm.m om.m mmmam mm.m ww.m Hm¢m on.m mm.m Hmdm w mH.m om.m mmdz HH.m mH.m mad: Ha.m mm.m ommo mumflmmumna pwogmauomxogH pan mmoum MOHgom How mm Hmmmm paw .xmmm .gmoz AH mamSquav conflnfiou mcoflmwmw mumthaumm mam .mumq .wanwm Hm>o mumflmonmsa poogoflHDQXDCH paw poocoflnmmxm How md Hmmmm Ugo rxmom .gooz m canoe a manna 50 2.7-r 32:31 2.6~r inexperienced therapists in 2.5% both sessions 2.4-- . . combined exper1enced therapists 2.3‘~ female male Clients Figure 5. Basal AB in both sessions combined broken down by sex of client and therapist experience (statistically significant interaction using ANOVA; p < .05) Thergpist Experience and Sex of Therapist In late therapy, all experienced therapists and inexperienced female therapists offer similar peak empathy, while inexperienced mglg therapists offer far lower peaks (see Table 6 and Figure 6). When senior staff alone are compared to all other experience levels (Analysis II), male senior staff offer quite high peak empathy, both inexperienced female thera— pists and female senior staff offer moderately high jpeak empathy, and inexperienced male therapists offer low peak empathy (p < .05). The 3-way interaction mentioned earlier on page 46 lalso has relevance here. No other statistically signifi- <:ant findings were generated regarding this interaction. 51 Table 6 Analysis of Variance of Peak AB in the Late Session by Sex of Client, Sex of Therapist, Experience of Therapist, and Client Group Status (Analysis I) Source of Significance Variation SS DF MS F of F Main Effects Sexcl .005 l .005 .055 -- Sexth .005 l .005 .054 -- Expth .333 1 .333 3.358 .074 Grst .273 l .273 2.751 -- 2-way Interactions sexcl sexth .602 l .602 6.071 .019 sexcl expth .058 l .058 .581 -- sexcl grst .292 1 .292 2.949 .093 sexth expth .437 l .437 4.408 .042a sexth grst .031 l .031 .314 —— expth grst .066 1 .066 .669 -- residual 2.874 29 .099 total 4.694 39 .120 aSee Figure 6. 3 8-~ r//////”/A experienced therapists Peak AE 3.7__ 1n the late 3 6~~ session ' 3.5-- inexperienced therapists female male Therapists Figure 6. Peak AB in the late session broken down by therapist experience and sex of therapist (sta- tistically significant interaction; p < .05) 52 Summary_of Hypothesis_2 Many results, both in Analysis I as well as in Analysis II, indicate support for Hypothesis 2 in the £333 session: (a) while the experienced therapists seeing group 2 clients offer mean AE levels similar to that offered by other therapists, they offer higher mean empathy to group 1 clients, (b) experienced mglg_thera- pists offer higher peak empathy than do the inexperienced male therapists, and a trend exists for the experienced in general to offer higher peak empathy than the less experienced therapists, (c) various correlations indicate fairly strong statistical relationships between mean and peak empathy and therapist experience. Thus, Hypothesis 2, though not confirmed for early sessions, is supported in many cases in the late session. Exceptions are that in the late session (1) experienced female therapists only offer similar (and not higher) peak empathy than do inexperienced female therapists, and (b) the inexperienced therapists are better at avoiding low basal empathy with group 2 clients and male clients than are the experienced. Even with the above exceptions in mind, overall partial acceptance of Hypothesis 2 seems both appropriate and reasonable. 53 Client Discomfort and Therapist Experience Hypothesis 2: Statement 3a: There is no interaction between client discomfort and therapist experience with regard to AE. Statement 3b: Experienced therapists provide higher AE to group 1 clients than to group 2 clients. Statement 3c: Inexperienced therapists provide lower AE to group 1 clients than to group 2 clients. Statement 3d: Experienced therapists will provide hi her levels of AE to group 1 clients than to group 2 clients, whereas inexperienced therapists will provide higher AE to group 2 clients than to group 1 clients. For early sessions, there is no difference between experience levels of therapist on mean, peak, or basal empathy with group 1 versus group 2 clients. For late sessions, statement 3d is essentially true for mean empathy, except that the mean empathy offered by experienced therapists to group 2 clients is the same as, and not higher than mean empathy offered by the inexperienced to such clients. Experienced therapists avoid low basal empathy in the late session with group 1 clients. What was 54 unexpected is that inexperienced therapists are able to avoid low basal empathy with group 2 clients while experienced therapists are not (p < .05). This inter— action is also statistically significant for basal empathy in both sessions combined (p = .01).4 Empathy Across Time Hypothesis 4: Accurate Empathy offered will be significantly higher in later therapy sessions than in earlier therapy sessions. Empathy Across Time—~C1ient Subjective Discomfort Main Effects There are no statistically significant differences in mean, peak, or basal empathy across time to clients with high and low subjective discomfort. When senior staff are compared to all other therapists combined (Analysis II), a strong trend (p = .058) suggests that group 1 clients receive increases in peak empathy across time, while group 2 clients receive decreases. Empathy across time—~Client discomfort and sex of client. Female clients from high and low subjective dis— comfort groups receive similar peak empathy across time, anxious—depressed male clients receive peak empathy 4See page 42 for further results. 55 increases across time, and other male clients receive peak empathy decreases across time (see Figure 7). +.31~ Group 1 +.2** Peak +.l‘t AE across 04~ ——————————————— no change time -.l‘” ...2 .1... -.3‘- Group 2 p —4 female male Clients Figure 7. Peak AB in late session as compared to early session broken down by sex of client and client group status (Analysis I) (p < .05) The above pattern is more marked for Analysis II (E < .01). There are no statistically significant inter- actions for mean or basal empathy across time. Empathy across time-~Client discomfort, therapist experience, and sex of therapist. Analysis II suggests a tendency (p < .10) for female senior staff and all male therapists to increase mean empathy across time to anxious—depressed clients, while a tendency for inexper- ienced female therapists to markedly decrease mean empathy across time to such clients. Furthermore, male senior staff and inexperienced female therapists increase mean empathy across time with clients low in subjective 56 discomfort, while inexperienced male therapists and female senior staff decrease mean empathy across time to such clients. There is a similar pattern (as the above) for peak empathy across time (Analysis II; p = .06). Empathy Across Time——Experience of TherapISt Mth Effects Experienced therapists have a tendency to increase peak empathy across time (p < .10). Senior staff, when compared to all other therapists combined, increase mean and peak empathy across time, while inexperienced therapists remain unchanged (Analysis II; p < .05). Empathy across time——Therapist experience and sex of client. There are no statistically significant inter— actions regarding empathy across time to male or female clients by experienced or inexperienced therapists. Analysis II, however, suggests that senior staff greatly increase peak empathy across time to female clients, while inexperienced therapists decrease peak empathy slightly across time to female clients. Furthermore, male clients receive essentially no change in peak empathy across time from senior staff and inexperienced therapists (see Figure 8). 57 Peak AE across time inexperienced therapists senior staff female male Clients Figure 8. Peak AE across time broken down by sex of client and therapist experience (Analysis II) (p < .05) Empathy across time-~Therapist experience and sex of therapist. Experienced male therapists and inexper- ienced female therapists increase peak empathy across time, while inexperienced male therapists decrease, and experienced female therapists remain essentially unchanged (see Figure 9). Male senior staff greatly increase peak empathy across time, while experienced and inexperienced female therapists, as well as inexperienced male therapists remain essentially unchanged (Analysis II; p < .05). Empathy Across Time--Sex of ClIént Main Effects There are no statistically significant main effects which involve mean, peak, or basal empathy across time to male or female clients. 58 experienced therapists Peak AE across time inexperienced therapists _,3—_ female therapists male therapists Figure 9. Peak AE across time broken down by sex of therapist and therapist experience (p < .05) Empathy across time—~Sex of client and sex of therapist. In terms of peak empathy, male therapists remain essentially unchanged across time with clients of either sex, as do female therapists with female clients. Female therapists tend to decrease peak empathy across time with male clients (p < .10). Thus, male therapists are consistent in their increases with male or female clients, while female therapists increase with female and decrease with male clients. Empathy Across Time-—Sex of TherapiSt Main Effects Male therapists increase basal empathy across time, while female therapists decrease basal empathy across time (see Figure 10). 59 +.37 Basal +'2- AE +.1_- across time O__ ___________________________ no change _.1-~ _.2 -- —'3_I- l female therapists male therapists Figure 10. Basal AE across time broken down by sex of therapist (p < .05) Summary of Hypothesis 4 It appears that the overly ambitious Hypothesis 4 is too broad to encompass the various possibilities that take place within therapeutic relationships. It is clear that empathy does not increase across time in all cases, but there are sufficient supportive results of the hypothesis to warrant partial confirmation. Sex of Client Hypothesis 5: Female clients will elicit higher AE across therapist experience levels than will male clients. Sex of Client Main Effects There exist no statistically significant findings regarding overall empathy differences between male and 60 female clients. This is true for mean, peak, or basal empathy, over early, late, or sessions combined. Sex of Client and Therapist Experience See page 48. Sex of Client and Sex of Therapist Male therapists offer moderately high peak empathy in the late session to clients of either sex, while female therapists offer very high peak empathy to female clients and low peak empathy to male clients (see Figure 11). 3.8‘* Peak 3.7" male therapists AB in 3 6‘- late ' female therapists session 3 5‘_ l A T 1 female clients male clients Figure 11. Peak AB in the late session broken down by sex of client and sex of therapist (p < .05) In addition, there is a strong trend which suggests that, in the piggy session, female clients receive moder— ately low basal empathy from male or female therapists. Male clients, on the other hand, receive low basal empathy from male therapists, and high basal empathy from female therapists (p_= .054). 61 Summary of Hypothesis 5 In general, female clients do ESE receive higher empathy than do male clients. In addition, female clients do not receive higher empathy from inexperienced therapists than from experienced therapists. Hypotheses 5 and 5b therefore must be rejected. Sex of Therapist Main Effects There is a weak trend which suggests that female therapists offer higher basal empathy than do male thera- pists during early sessions (p = .082). Interestingly enough, during the late session, there is a trend (p_= .091) that suggests the reverse (i.e. that male therapists offer higher basal empathy than female therapists). CHAPTER IV DISCUSSION General Considerations Hindsight suggests that the hypotheses were stated too broadly. Nevertheless, the results suggest interesting implications. In general, empathy levels present during saggy sessions are similar for all therapist-client combinations. In contrast, very dis- tinct differences in empathy emerge when the late session is examined. Combining early and late ratings generally results in a masking of differences that occur during the late session of therapy. Psychotherapy may consist of several stages rather than a continuum. One may speculate that therapists ini— tially pass through a "warm—up" period—~a time when they receive the client's statements with reasonable (but not too high) understanding. Because therapists do not know the client well yet, perhaps they (the therapists) remain in "low gear" rather than attempting too much too soon. More specifically, if one conceptualizes a two— stage model of the psychotherapeutic process (early and 62 63 late), the early stage is superseded by a new level of interaction. In some instances, empathy increases from early to late sessions, while in others empathy decreases or remains unchanged. Support for a two—stage model is suggested by the relative independence of early and late empathy ratings. One cannot tell what level of empathy will be present in late therapy from levels attained during early therapy. Whether empathy stabilized is not clear though. Additional empathy ratings to that made of the late session (3/4 mark of therapy) would have highlighted the point at which therapists offered varying amounts of empathy. The session in which empathy became consistent would also have been clarified. Because differences in empathy offered in the early session were minimal, the late session will be the focus of this discussion unless otherwise stated. High Versus Low Client Subjective DiScomfort (HypotheSis 1) Client Supjective Discomfort Main Effects Anxious-depressed clients in general tend to receive empathy levels similar to or higher than levels of other clients. This supports earlier notions and the basic tenet of Hypothesis 1 that clients bring some char— acteristic or attitude to therapy which increases the likelihood of empathic behavior on the part of the 64 therapist. Therapists, while not responding with a high average empathy level, do generate quite high pggx empathy with anxious-depressed clients. It may be that these peaks are in response to particularly strong expressions of discomfort. Clients reporting little subjective dis- comfort seem unable to elicit such intense therapist efforts in regard to communicated understanding. Client Subjective Discomfort and Experience of Therapist Clients who seek help but who report little sub- jective discomfort seem to receive similar mean empathy levels from both experienced and inexperienced therapists, and surprisingly, higher basal empathy levels from inexperienced therapists. Clients who are anxious- depressed seem to receive far higher levels of empathy from the experienced therapists. This presumably occurs because the experienced therapists sense the clients' needs 329 are able to deal effectively with their own anxiety. So, when clients have much subjective discom— fort, experienced therapists increase their empathy. Inexperienced therapists, possibly made more anxious by client distress, decrease empathy to anxious-depressed clients. One could argue that the inexperienced therapists avoid low basal levels with low subjective discomfort clients because they are so pleased to be with clients 65 who are help-seeking yet not forcing them to deal too much with their unresolved feelings regarding the client's dependency and their own helplessness. On the other hand, the inexperienced therapists may either have a greater need to get clients involved due to their (the therapists') uncertainty, or they do not sense the factors which make more experienced therapists cautious in emitting empathy with these low subjective discomfort clients. Client Supjective Discomfort and Sex of Client Anxious-depressed mgigs tend to receive higher levels of peak empathy than their less distressed counter— parts (Analysis II). Since males culturally do not show feeling as frequently as do females, perhaps male clients who are high in subjective discomfort are perceived to be in more pain or need (because they differ so much from the male stereotype). Thus, client subjective discomfort in tandem with maleness elicits high peak levels of empathy. In contrast, males without this reported discom— fort receive 1ow empathy levels. Therapists may sense that these males are "aggressive—resistive," thereby making it difficult to offer empathy to them. "Extinc- tion" from the learning theory framework may in fact be what is operating here. 66 Female clients with high subjective discomfort receive only slightly higher peak empathy than female clients with lower discomfort. Thus, in general, one need not know how anxious—depressed a female client is to surmise the peak empathy she will receive. The sub- jective discomfort breakdown seems important mostly for mgigg. It is not clear why female clients failed to stimulate such dramatic differences. Client Subjective Discomfort and Sex of TherapISt Male and female therapists offer similar empathy levels to clients of both high and low subjective discom— fort. Experience of Therapist (Hypothesis 2) Experience of Therapist Main Effects In general, therapist experience does seem to be related to empathy offered in the late session of therapy. There is a statistically significant correlation between experience and peak empathy, as well as a trend toward experience level's relationship with mean empathy. The therapist experience—peak empathy correlation is corrob— orated by a trend (ANOVA) for more experienced therapists to offer higher peak empathy in the late session. The data thus support previous research on that topic. 67 Therapist Experience and Sex of Client Female clients seem to have no advantages over male clients with regard to eliciting mean or peak empathy, no matter what the experience level of the therapist. When combining early and late sessions, how- ever, experience does seem to be an asset in avoiding low basal empathy with females, but becomes a liability in terms of reduced basal empathy with male clients. It may be though that experienced therapists sense more accurately what clients are seeking. Alexander (1967) cites Apfelbaum's (1958) finding that males expect a "critic," and females a "permissive listener" as a therapist. Client expectations then may be different. In order to engage male clients (and particularly males reporting little distress), experienced therapists may offer lower basal empathy believing that too high empathy will create dissonance with what male clients expect. Therapist Experience and Sex of TherapISt Female therapists offer similar levels of empathy whether experienced or not. Experience is a significant factor in terms of peak empathy offered by male therapists. Experienced male therapists (senior staff and second-year interns combined) achieve as high peak empathy as do all female therapists, while inexperienced male therapists fall below those levels. Male senior staff alone offer 68 high peak empathy (even higher than all female therapists), but inexperienced male therapists offer low levels of peak empathy. Sex roles offer especially viable explanations for many of the differences found related to sex of therapist (and also sex of client). It is suggested here that the process of becoming a therapist may be different for males and females. This possibility will be discussed more fully later (see section on "Training Implications"). Client Subjective Discomfort and Therapist ExperIence THypothesis 3) Exploratory statement 3d maintains that exper- ienced therapists will provide higher empathy to anxious- depressed clients than to other clients, while inexper- ienced therapists will provide higher empathy to clients reporting little subjective discomfort. This is essen- tially borne out by the data. One may speculate that clients reporting low anxiety~depression were not seeking high levels of empathy, and thus both the experienced and inexperienced therapists could comply. When greater empathy was "solicited" by high subjective discomfort clients, only the experienced could generate more empathy and respond adequately. Parallel findings occur in basal empathy scores; in this instance the inexperienced therapists avoid low 69 basal empathy with low subjective discomfort clients while experienced therapists do not. With this added infor- mation, one might speculate that experienced therapists consciously reduce empathy to low subjective discomfort clients in the belief that the lower empathy style will be more effective. In summary, to inexperienced therapists, clients reporting little subjective discomfort may appear to be easier to work with than anxious-depressed clients. The inexperienced typically respond more empathically to low distress clients, while the experienced are not turned aside by higher anxiety of the client.1 Empathy Across Time (Hypothesis 4) It is apparent that empathy typically increases across time, yet this is not a universally occurring phe- nomenon. There are many therapist-client combinations in which empathy either remains unchanged, or decreases across time. Empathy Across Time--Client Subjective Discomfort Globally speaking, anxious-depressed clients do not receive any more or less empathy across time than do clients with little subjective discomfort. The exception 1See page 64 for added discussion of the therapist experience-client subjective discomfort interaction. 70 is that there is a trend for clients with high discomfort to receive increased peak empathy across time. Empathy across time—~Client subjective discomfort and sex of client. Combining both high and low discomfort clients seems to dilute differences that occur when spe- cific breakdowns are examined. For example, when peak empathy across time is analyzed for all male clients, the result is no change. However, anxious-depressed males receive large increases of peak empathy across time, while low subjective discomfort males receive large decreases of peak empathy across time. Clearly, knowing whether a male client reports high subjective discomfort is critical when predicting peak empathy across time. Again, I mention that it is a cultural stereotype that males do not show feeling as easily as do females. Male clients who report high anxiety-depression have deviated from this stereotype. To the extent that therapists share this View of men, they may perceive these male clients to be in much distress, and needy of especially high levels of empathy. In contrast, female clients receive little or no increases of empathy across time regardless of subjective discomfort levels. Females who are anxious-depressed do not receive the empathy increases that males with such discomfort do. Perhaps therapists sense that the report— ing of discomfort for females is in keeping with culturally 71 shared views of females' expression of feeling, and there- fore see no cause for increased empathy. Females not reporting high subjective discomfort are receiving peak empathy across time akin to that accorded other female clients (i.e. no change). Thera— pists are maintaining peak empathy with these females even though substantial client distress is absent. Female clients reporting little subjective discomfort may be receiving a benefit from the sex stereotype, for men with similar low levels of discomfort receive reduced peak empathy across time. Empathy Across Time--Experience of Therapist When senior staff and second-year interns are compared to first-year interns and practicum students, there is a trend for the former group to increase peak empathy across time while for the latter to decrease peak empathy across time. Senior staff (Analysis II) show marked increases in mean and peak across time when com- pared to all other experience levels combined. This is probably due to the experienced therapists' effectively utilizing what new information becomes available as they come to know their clients. The inexperienced are apparently not able to use these cues as efficiently. Empathy across time--Experience of therapist and sex of client. All combinations of therapist experience 72 by sex of client receive essentially little change of peak empathy across time, except that senior staff greatly increase in this regard when seeing female clients. Females expect a permissive listener (Apfelbaum, 1958), and experienced therapists may just be better able to perceive and deliver what the client is hoping for. Empathy across time-—Experience of therapist and sex of therapist. There is a trend for experienced male therapists to increase mggp_empathy across time, and for experienced female therapists to decrease mean empathy across time. In addition, experienced male and inexper— ienced female therapists increase pggx empathy across time, while inexperienced male therapists decrease peak empathy across time. Male senior staff, when compared with all other experience of therapist by sex of therapist combinations, markedly increase peak empathy across time. Male senior staff may begin cautiously to determine what needs to be empathized with, and communicate this added understanding only later in therapy. Empathy Across Time—~Sex of Client and Sex ofiTherapist A trend suggests that there were slight peak empathy increases across time for all sex of client by sex of therapist combinations, except for a decrease 73 when female therapists see male clients. This finding will be discussed in the section "Same—Sex Versus Cross- Sex Dyads." Sex of Client (Hypotheses 5 and 5b) See sections "Client Subjective Discomfort and Sex of Client" and "Therapist Experience and Sex of Client" for discussion. Issues Training Implications General considerations. Fischer (1975) recently found that a group that was given training in core con— ditions moved 2.7 steps higher on Truax and Carkhuff's nine—point empathy scale. This movement was significantly different from that of his contrast and control groups. Mullen (1969) has already shown that empathy is trainable. In his study, empathy changed with increased training and/or experience. If this last statement is a valid one, why were the experienced therapists app higher on empathy in all cases? Dietzel (1971) discovered that successful psycho— therapy was marked by the therapist's ability to depart from the original complementarity he held with a particular client. This was especially critical during the middle stage of therapy. Is it possible that experienced therapists sense that empathy has utility only at certain 74 stages of therapy, and emphasize other facets of their behavior as the need arises? Perhaps "the more the better" does not apply to empathy in all cases. An additional explanation may be that inexperience (or something about the inexperienced) may be an asset with some clients. Perhaps transference difficulties, or even more overt authority problems are disarmed more easily by inexperienced therapists. Practicum students and first-year interns. In several cases, practicum students offer very high empathy levels. The practicum students may offer very spontaneous remarks which, springing from good intuition, prove to be accurate. Something occurs in the first year of intern- ship, however, which causes upheavals. The first-year intern is not empathic with anxious-depressed clients, perhaps reflecting his or her own anxieties (especially since this is his or her first increased case load). It may be that the first-year intern has some training but not enough to put into effect the empathic response intended. His or her responses may be stilted and reflect a using of technique (e.g. "I hear you") without true understanding. First-year interns may be suffering from what Abeles (1963) called "affective lag." He theorized that "the ability to be aware of one's own affect pro— gresses at a much more rapid pace than one's ability to respond affectively to others" (p. 4). In the process 75 of sorting all that they are learning, first-year interns may temporarily lose effectiveness. First-year interns, who give lower mean empathy to anxious-depressed clients, may be particularly under pressure themselves. First-year interns are expected to have some expertise, but really have bare essentials in therapeutic skills. They may feel threatened by these demanding clients, but do not have the time to delve into any one client, either in terms of psychic energy or in supervision. Also, as compared to practicum students, ILL-I‘ll the first-year interns may hold very different expectations of themselves (i.e. "I am no longer a beginner"). Hartzell (1967) found that interns had the highest need to "nurture," with practicum students next, and staff last. So although interns have a high need to nurture, perhaps anxious-depressed clients never give the interns the feeling that they (the clients) received nurturance. That is, the therapists nurture as long as the client "requests" it, but client demands are perceived as insatiable. On the other hand, perhaps the clients per- ceive the interns as trying too hard, or offering empathy out of the therapist's own needs rather than for the client. Secondjyear interns. By the second year of internship, therapists have apparently gained much in terms of being able to deal with anxious—depressed 76 clients. Whereas first—year interns may have anxiety which interferes with their desire to be empathic, perhaps second—year interns have mastered their own discomfort with anxious-depressed clients. This mastery permits them to offer higher amounts of empathy than their less experienced fellow interns. Second—year interns are not especially empathic with clients reporting relatively low subjective discomfort however. My hunch is that the therapists' anger, rather than anxiety experienced toward these low subjective dis— wl comfort clients, interferes with their ability to empa- thize. "Why are these people here? Are they motivated to change? Why aren't they anxious?" may be some questions that arise for second-year interns as they interact with these clients. Senior staff. Senior staff are more likely to be tolerant of a wider range of client behaviors; thus they should be able to treat clients more uniformly. In addition, senior staff may have resolved their own anxiety BBS anger as potential interferences in the therapeutic encounter. In fact experienced staff members offer similar levels of mean empathy in the late session to BEER high and low subjective discomfort clients. Regarding this issue, senior staff have stabilized their empathy. 77 Senior staff increase mean and peak empathy across time, while the inexperienced therapists do not. The experienced apparently utilize their increasing knowledge of the client to advantage by communicating understanding. The inexperienced, not having honed skill at evaluating client status and difficulty, rarely add to their empathic ability with a particular client. Becoming a therapist--Males versus females. The process of becoming a therapist may differ for male and female therapists. Male therapists must learn to move ! away from the stereotypic male role of avoiding feelings. Inexperienced male therapists cannot respond empathically to anxious-depressed clients, whereas inexperienced female therapists can. The anxious—depressed client may force the inexperienced male therapist to examine his own feelings and vulnerability. By the time the male has become experienced, he does offer high empathy to the anxious- depressed client. Becoming experienced for female thera- pists may entail moving away somewhat from their natural tendency to be empathic, and more toward confronting, assertive components of therapy. It is speculated that graduate training is more important in the development of male therapists than of female therapists, at least when examining the peaks of empathy which therapists attain. Female therapists, whether inexperienced or experienced, offer high peak 78 empathy, as do experienced male therapists. Inexperienced male therapists offer far lower peak empathy than do other therapists. If one considers the traditional female stereotype of more openness to feelings, the above becomes understandable. Inexperienced female therapists offer initially high levels of empathy, and experienced females maintain but do not exceed those levels. Inexperienced male therapists initially offer relatively low peak empathy, and only attain high peak empathy when they become exper— ienced. Training then seems to increase a male therapist's ability to achieve high peaks of empathy. In summary. It seems likely that there are varying issues with which developing therapists must deal, and that these issues emerge at different stages of training. It is very clear that interns work quite differently with varying client types, and counseling centers must be attuned to these differences. These results also throw some questioning light on earlier research where first— and second-year interns were considered a homogeneous group. Certainly in the future, researchers must be more cognizant of potential differences between more and less experienced interns. Various studies have found therapist experience to be the primary variable, or at least an important variable in psychotherapy (Mullen & Abeles, 1971; Hart— zell, 1967; Bienenfeld, 1975; Scher, 1975). The results 79 of this project suggest that therapist experience is indeed an important factor when examining the level of Accurate Empathy achieved. Training programs should con- tinue to emphasize empathy skills as an asset in a thera- pist's development. Same-Sex Versus Cross-Sex Dyads Same—sex pairs seem to be equal to or better than cross-sex pairs in terms of peak empathy generated. Very high peaks of empathy are offered when female clients are seen by female therapists, whereas much lower peak empathy is offered when male clients are seen by female therapists. Male therapists offer similar peak empathy to clients of both sexes. Perhaps the male therapist has had far more contact with female clients than the female therapist has had with male clients. (In college, female clients are more likely than males to seek help, and are more likely to be seen by a male therapist than by a female therapist.) Cartwright and Lerner (1963) found that therapists obtained significantly higher empathy scores with patients of the opposite sex than with those of the same sex, but this difference disappeared by the end of therapy because empathy for clients of like sex increased significantly over the course of treatment. In contrast, my results indicate little difference between same—sex and 80 opposite—sex dyads in ggxly_therapy, but some support for same-sex pairs in $333 therapy. Riess (in Bergin & Suinn, 1975) found that patients remain in psychotherapy longer with female therapists, and female-female dyads continue treatment longer than male-male dyads. His results can be par- tially explained by my findings that female clients receive the highest peak empathy during the late session from female therapists. To the extent that female clients perceive this high degree of communicated understanding, I they will tend to remain in therapy longer. Bienenfeld (1975) indicated that though female clients received more actual empathy than male clients, they also received more variable empathy than did males (actual empathy in general was negatively correlated with variability of empathy). Summarizing previous research, she maintained that . . . women may prove to be less consistent (i.e. their behavior may be more situationally determined than the behavior of men) in the level of therapeutic conditions they offer in a counseling situation. (p. 15) The combination therefore of female client and female therapist may result in increased empathy variability as a by-product of high actual empathy. Are therapists acting as counter—culture change agents, and responding to those who attempt to vary from societal norms? Bienenfeld (1975) mentioned that 81 therapists of both sexes may be cross—sex identified, and hence more likely to emphasize movement away from the traditional sex roles. This possibility seems tenable when Cartwright and Lerner's (1963) findings are con- sidered. They maintained that . . . therapists had more initial difficulty under- standing like sex clients than clients of the opposite sex, but this handicap was overcome with time. (p. 142) They suggested that assumptions of similarity are lg§§_ likely to occur with opposite sex clients, leaving the therapists freer from a projective set. Meltzoff and Kornreich (1970) stated that there was "no clear basis for preferential assignment of a patient of either sex to a therapist of either sex." However, female therapists do offer high empathy with female clients. Male therapists, though offering lower basal empathy with male clients than female clients in early therapy, increase both peak and basal empathy across time. In late therapy then, male therapists offer similar basal empathy, and higher peak empathy to male clients than do female therapists with male clients. Though there were no consistent sex of client by sex of therapist mean empathy differences, to the extent of empathy's importance to clinical outcome, the above is evidence in support of same-sex pairing. CHAPTER V SUMMARY Empathy has been widely studied in order to assess its contribution to the process of psychotherapy. This study's primary purpose was to evaluate the degree to which empathic understanding occurred with two differing client "types." One group, a high subjective discomfort group (elevated MMPI depression and psychasthenia scale scores), has been shown by previous research to be a mean— ingful typology for college students. The second client group, a low subjective discomfort group, consisted of those with significantly lower scores on the MMPI depres— sion and psychasthenia scales. Differences between the two groups regarding age of client, verbal and total scores on a standard college entrance examination, and all other MMPI validity and clinical scales were essen— tially nonsignificant. The completed psychotherapies of 40 college student clients were examined using Carkhuff's (1969) Empathic Understanding in Interpersonal Processes Scale. 82 i 83 The following questions were considered: Do anxious-depressed clients receive more empathy than other clients, do experienced therapists offer more empathy than inexperienced therapists, do therapists increase empathy offered across time, and is there a client sex difference in empathy received? Discussion centered not only on the above, but also on differences in same—sex versus cross—sex dyads in relation to empathy levels generated. While the hypotheses were stated too broadly, the results suggest support for the notion that therapists offer differential amounts of empathy with clients of differing characteristics. The empathy level in any therapeutic encounter depends on the sex of client or therapist, experience level of therapist, and client type. Findings include the following: (a) experienced therapists offer higher empathy in general, but particu— larly with clients who are reporting high subjective dis- comfort; (b) empathy ratings from late sessions of therapy provide far more statistically significant results than do those of early sessions; (c) there is qualified support that anxious-depressed clients receive higher empathy than other clients; (d) female clients as a group do not receive higher empathy than do male clients; (e) same—sex dyads are more likely to achieve high peaks of empathy than are opposite-sex dyads; (f) experienced therapists tend to increase peak empathy 84 across time; (g) anxious-depressed clients tend to receive higher peak empathy across time, while other clients tend to receive lower peak empathy across time; (h) anxious-depressed male clients receive increases in peak empathy across.time, while other male clients receive decreases in peak empathy across time; and (i) senior staff greatly with female clients, but Overall, clients levels of empathy. Peak session seem to generate increase peak empathy across time not with male clients. in this study receive acceptable empathy scores in the late the most significant differences. Low and peak empathy ratings were significantly correlated for early and late sessions, as well as combined sessions. It appears that therapists offer high empathy levels while avoiding low levels. Implications of these findings suggest the need to control for client personality characteristics even in relatively homogeneous college student client populations. APPENDICES APPENDIX A INFORMATION ON SELECTED MMPI CLINICAL SCALES APPENDIX A INFORMATION ON SELECTED MMPI CLINICAL SCALES The D scale (scale 2) was designed to measure "pessimism of outlook on life and the future, feelings of hopelessness or worthlessness, slowing of thought and action, and . . . preoccupation with death and suicide" (Dahlstrom, Welsh, & Dahlstrom, 1972, p. 184). The scale's reliability, as reported by Hathaway and McKinley (l956),was equal to .77 i .04. Many offered various descriptions of depression and its behavioral signs (Zubin, Salzinger, Fleiss, Gur- land, Spitzer, Endicott, & Sutton, 1975; Wechsler, Grosser, & Busfield, 1963; Beck, 1973), but all descriptions can be summarized by the five identified factor clusters offered in An MMPI Handbook, Volume I (Dahlstrom et al., 1972): subjective depression, physical malfunctioning, mental dullness, psychomotor retardation, and brooding (p. 407). Beck (1973) called depression a "primary mood disorder" or "affective disorder," and depicted the emotional, cognitive, motivational, and physical mani— festations that may accompany depression. 85 86 The Pt scale (scale 7) was designed to help evaluate the obsessive-compulsive syndrome, but also taps "abnormal fears, worrying, difficulties in concen- trating, guilt feelings, and excessive vacillation in making decisions" (Dahlstrom et al., 1972, p. 211). The items cover "anxiety and dread, low self-confidence, doubts about one's competence, undue sensitivity, moodi- ness, and immobilization." Dahlstrom et a1. (1972) reported a correlation of .26 between scales 2 and 7, with 13 items overlapping both scales. Barron (1956), however, cited several sources which found correlations from .41 to .57 depend- ing on which college group was used. Scale 7 correlated .53 with the F scale, but only -.17 with K. Though Pa is essentially independent from D and Pt, Pa does overlap with scale 2 on two items, and with scale 7 on four items. The difference on Pa between groups 1 and 2 in this study could have occurred had group 1 members answered two more items (on the average) in the critical direction. "Because of this overlap in the component items, the basic scales possess varying amounts of experimental dependence" (Dahlstrom et al., 1972). For example, there are 32 depressive affect statements in the MMPI: 13% come from the D scale, 10% from Pa, 17% from Pt, 12% from Sc, and 4% from Si. APPENDIX B CHARACTERISTICS OF D—PT CLIENTS APPENDIX B CHARACTERISTICS OF D-PT CLIENTS Depression and anxiety frequently occur together clinically and "both states seem to be triggered by failure-related threats to self-esteem" (Becker, 1974). He even suggested that they be combined into a category called "affective states." Anxiety may just be an alternative way of experi- encing depression. Zubin and Fleiss (1971) reported a cross-national study which found that depressive neurotics and anxiety neurotics were no different with regard to depression. Depression and anxiety factors' correlation was equal to .48. Costello (1970) discussed factor analyses of psychopathology, and indicated that a con- sistent result was the combination of depression and anxiety into a single factor. Thus, examining them together has precedence and can be done meaningfully. Hathaway and Meehl (1956) suggested that the presenting complaints of D-Pt clients were "depression, with tenseness and nervousness as frequent accompaniments." Many suffer from anxiety, insomnia, and undue sensitive- ness. The men, according to Guthrie (1956), also showed 87 88 rigidity and excessive worrying. Drake (1956) found that the "27" profile tended to be elevated among college counselees, and seemed to be related to home conflicts. Plyler (1965) compared students seeking help with vocational, educational, or emotional problems, and found that for the emotional problem group: D was the highest scale 12.9% of the time, Pt the highest 20.4% of the time, D was second highest 10.9%, Pt the second highest 14.9%, and D EEE.Pt were the two highest 10.2% of the time. The most typical two-point patterns for the "emotional" group were Pt-Sc, D-Pt, Sc-Pt, Sc-Ma, and Ma-Mf. Plyler stated that D-Pt with students: . . . may be associated with anxiety with depressive elements possibly included. This particular pattern is often considered an indication of readiness for counseling and consequently may be a good prognostic sign for college students even though it may have more serious implications such as suicide risk in psychiatric groups. (Plyler, 1965, p. 58) Bergin and Solomon (1970) described D and Pt as indicators of subjective discomfort, and it is these two scales which my study has used to discriminate the client groups. Canter (1960) maintained that high D and Pt scores were frequently found with those 83 having loss of self-confidence independent of specific diagnoses. He devised a Morale Loss Scale (ML) which essentially had all the items from the D, Pt, and K scales. He was able to differentiate among severely depressed patients with 89 suicidal attempts, severely depressed patients, and normals with this scale, and thus suggested D and Pt as indicators of depression and low self-esteem. Most recently, Strupp and Bloxom (1975) tested the entire freshmen class of a university, and put all men who had D, Pt, and Si above T = 60 on the MMPI into a group. They randomly selected a comparison group from the remainder of the class. Results seemed to indicate that these men were more likely to have vocational and emotional problems in college, as well as at an 8 1/2 year follow-up. The 27 configuration then, they maintain, is "indicative of a genuine clinical problem with demon- strable and enduring negative intrapsychic and behavioral consequences." These men also considered themselves more troubled and sought more professional help than the random group. While Strupp and Bloxom may have discovered that high 27 men were help-seeking and had some long-standing problem, the question emerges, "compared to whom?" They told us very little about their "random" group. The study would have had far more weight had the comparison group been of individuals who had two other scales above T = 60 on the MMPI. At least then, overall elevations would have been similar. Their study essentially informed us that those reporting anxiety and depression would seek help more 90 often than a random group, but this is not new infor- mation. We would expect these men to seek help. This group is aware of their discomfort unlike other perhaps maladjusted yet defensive 55. That the high 27 group considered themselves more troubled is consistent with low self-esteem and morale loss which accompany depression and anxiety. Furthermore, the MMPIS in question were taken as the 85 entered college. We know little about how the pro- files were at the time when some 83 sought help. Were they still high on scales 2 and 7? In addition, might not appropriate help-seeking, as well as awareness of conflict, be considered a strength in many cases? My main objection to their study then is their failure to focus sufficient attention on the "random" group. We need to know who the members in the group are. We also need the Means for the entire class on D and Pt, as well as on the other scales. Did they control for validity scale variations? The results as they stand have only marginal utility. My study has largely dealt with the main objection articulated above because BEER group 1 and 2 individuals sought help from the counseling center, and differences on most clinical scales were nonsignificant (except for the selected ones). Strupp and Bloxom maintained that: 91 . . . regardless of measurable "disturbance," a person who sees himself in need of professional help and who has proceeded to define himself as a patient is a very different person from his non- patient counterpart. This difference in self— definition is bound to have important consequences for psychotherapy outcome studies. . . . (Strupp & Bloxom, 1975, p. 236) - The "illness" then may be in the self-perception. This commonality among individuals who perceive themselves as in need is a further equating element to the similarity of groups 1 and 2 in my study. Merrill and Heathers (1956) reported significant negative correlations between D and the EPPS needs of Exhibition and Dominance, but positive correlations with Succorance and Abasement. Pt was positively correlated with Succorance and Abasement, but negatively correlated with Deference, Dominance, and Endurance. Thus both D and Pt were directly related to expressed needs of Succorance and Abasement. APPENDIX C RULES FOR RATING ACCURATE EMPATHY IN SPECIAL CASES APPENDIX C RULES FOR RATING ACCURATE EMPATHY IN SPECIAL CASES "uh-huhs" were not rated. If a therapist response follows another therapist response after a silence, two ratings were made. Ratings segments always start and finish with client statements. Appropriate questions of clarification by therapists were standardly rated level 3 (i.e. questions which did not "take away" from client affect). Level 3 rating does not have to include a feeling word if the therapist response summarizes the content and meaning of the client. Raters may go back and listen to tape segments more than once if they feel they can hear more data. A therapist statement interrupted by a client "uh-huh" is also rated once as long as the therapist is con- tinuing a thought (i.e. client saying "I'm listening"). Therapist response without client reaction (at end of segment) will not be rated. When the therapist expresses his feelings, generally a level 3 rating (unless significantly adds or sub— tracts from on-going material). 92 93 «H _:_____,-___.__._.._..- fi. APPENDIX D RATER RELIABILITIES APPENDIX D RATER RELIABILITIES Ratable responses per tape segment The lowest number of responses per 6.125 (mean) = 6 (median) segment was two, and the highest number of responses per segment was fifteen. A. Training Tape Segments .95 (mean .85 (peak .73 (mean .47 (peak .90 (mean .71 (peak QWnwaI-J IHHHHU1UHH empathy empathy empathy empathy empathy empathy B. Research Tape Segments .44 (mean .80 (mean 1 2 3 .96 (mean IHHHH "ll" empathy empathy empathy on on on on on on on on on segments segments segments segments segments segments segments segments segments 10-17) 10-17) 18-25) 18-25) 26-35) 26-35) 1-10) 20-29) 40-49) The above three reliabilities were sample relia- bilities for the actual eighty tape segments used in this project. A 26-segment estimate of the total reliability was then derived (every third segment starting at segment 3 and ending at segment 78). This was a reasonable approach since the tape segments were placed on the 94 95 master tape in a randomized order. The reliability estimate was very high for mean and basal empathy, and was moderately high for peak empathy: .88 (mean empathy on 26 segments) .48 (peak empathy on 26 segments) 4 5 6 .92 (basal empathy on 26 segments) IHHHH HIIH Because of the rather low peak empathy reliability estimate, further reliability checks were made: 7. £_= .72 (peak empathy on 26 segments starting at segment 2) 8. r = .70 (peak empathy on 27 segments — starting at segment 1) 9. £_= .61 (peak empathy for all research tape segments) __‘.. '_ .—.----‘ __ ,1 _-H. 1 —- APPENDIX E MMPI SCALED SCORES FOR CLIENTS OF BOTH GROUPS 7 Group 1 Earnesf Margie Kate Ben Sam Marilyn Date Ross Jean Mary (a) Coffee Tom Diane Richard Jerry Key David Ann Mary (b) Day Peggy Barbara Leon Rollo Abby Edith Mr. E Beth Shirley Elaine Gary Joan Henry Jim Sandy Mr. C Trudy Wendy Rita 46 2‘66 50 46 55 46 56 6O 50 4o 55 4o 45 40 4o 44 44 APPENDIX B Table Eel MMPI SCALED SCORES FOR CLIENTS OF BOTH GROUPS BE 42 54 56 65 54- 46 46 67 52 59 6O 59 44 52 62 54 54 54 48 4e 96 3g 60 71 6O 6O 74 9O 51 67 74 64 68 55 57 55 6O 55 79 64 67 52 Pd 58 86 76 55 48 76 65 48 64 55 5o 64 60 6O 75 50 57 67 55 69 Mf 69 45 48 78 57 95 113 75 52 59 59 69 41 57 1 .L 55 67 ll 5 50 54 Mf 8O 47 55 61 57 5Q 57 65 49 59 34 55 45 55 72 45 55 28 55 49 33 59 62 55 62 65 75 62 59 7O 59 56 55 7O 59 67 50 62 56 62 50 Pa 62 75 6? 55 SO 47 48 EB 69 76 68 6O 75 65 68 87 79 ~5 71 71 66 7.1 95 76 64 61 75 74 Pt 62 6] 55 56 66 48 40 60 6O 61 58 62 66 46 6O 48 45 66 55 68 s9 61 75 65 44 76 71 58 76 67 bl 79 75 R7 7'1 76 66 75 61 57 70 s3 65 75 58 55 50 4o, 57 75 65 51. 87 55 67 57 78 61 50 69 57 69 54 54 69 44 49 48 36 65 42 52 50 L). 5 44 75 42 62 66 45 APPENDIX F CLIENT CODE NAMES AND THE SESSIONS FROM WHICH THE TWO SEGMENTS WERE TAKEN 9.....r9 22.}. Earnest Margie Kate Ben Sam Marilyn Date Ross Jean Mary (a) Coffee Tom Diane Richard Jerry Key David Ann Mary (b) Day CLIENT CODE NAMES AND THE SESSIONS FROM WHICH THE TWO SEGMENTS WERE TAKEN sessions and and and and and and and and and and and and and and and and and and and and HHNHHNNJ-TNWWNNNONHWWWN Hummemt—Jmfl U1 (m # \NW 0‘ J? NUlU'I EURO APPENDIX F Grou9_§ Hans Peggy Barbara Leon Rollo Abby Edith Mr. E Beth Shirley Elaine Gary Joan Henry Jim Sandy Mr. C Trudy Wendy Rita sessions 14. HNHWOWPFNHWNHPNWHWNW and and and and and and and and and 15 10 5 14 5 10 6 12 4 (early and and and and and and and and and and 14 and late) APPENDIX G FOR EACH CLIENT: SEX OF CLIENT, EXPERIENCE LEVEL OF HIS OR HER THERAPIST, SEX OF THERAPIST, AS WELL AS MEAN, PEAK, AND BASAL EMPATHY OVER EARLY, LATE, AND SESSIONS COMBINED APPENDIX G Table G—l SEX OF CLIENT, EXPERIENCE LEVEL OF HIS OR HER THERAPIST, SEX OF THERAPIST, AS WELL AS MEAN, PEAK, AND BASAL EMPATHY OVER EARLY, LATE, AND SESSIONS COMBINED FOR EACH CLIENT: 75 mm Ln Ln In men Lmn NL\ NM N L\InN N[\[\ NNLn MHMMKOBMHOMBUJNCDMHHN 0,-00 00 00 at? I¢ H MNNNNHNNNMNHNNNNMMM N Immmn In Ln mm In MMNBNBOMMOBOEEONBQONOM o o o o o HMMHMHMHNNNMNNNNMNMMM .125 H Ln Ln In EMOOOOOBOOMQQMQOBQONO HHMMHMNHMMHMNHMMNMMMM In In men N NIn (\N In BN5 F11 mamoxoonomxoommmumcoaoo m°m¢m¢mmm¢mm¢mmmmmm¢m N m U\ U\L{\ Lf\ m fiNOBONOOBBONMMfiOOOO 0’ 0 o ‘2 m¢M¢¢M#M¢#MM¢MMMM¢¢#¢ H InLn Ln In InI-nln WBMOONNOMOOMOBMMBOfiNB Q-d Md’Md’Mthd'MMd'MMMMMMd'M InI-n InI-n LnI-n Lmn LnNgNHNOMHQMGQ-IBBQHBM £73 MN HOKOMNBNd'BHHHO\d’\O\O oo 0 000,000... .< Z MMMN MMN MMN MN‘ N MMMN MMM NInCOMO\MO\C\\OHd’H¢Md'L\HE\NOCD WBMC‘MQHMMHOMMWHHNmml-nm O O O O O O O O O O O O O O O O O O <: 2: MN MN MN MMMMMN MMMN MMM H MH Nmoomoxcotmn CD[\M H <1) WCNMLnMd'meMMNM OHHON m ZNMMNMNNMMNMMNMMMMMMM J: «P . xzzmzmzzzmmzzzmmzzzmz I2 I2 11 8 Il I2 S S S S I1 I2 S I2 Il I2 I2 8 zmmztzmmzmmmzmzzmzmmm sexcl ex th se P+ P HI-D ‘3 '6 mm >3 ID H 0-H H 0 06>; 'U SEMO HQMGE‘H GEL! H O gaucauumm H8m0H>~a>G >3 $466 wommmommoow-Hommg GS DEEMQWZQNWZOEQMWMQ EC: 98 APPENDIX GanContinued NI ‘6 >3 :3 0 gm mg 50>: .—-| . 5013,02th $19-th? 'U (3566 a NQOHQH Ova-4669 CBS 035343 Q) GOD'UHID—CHCUOG-HCUH m'v-i (D mmqmCHDnJ O O O 0 ~. 0 O O O O O O O O O O O O O NNNNNMMMNNMHNNMMM.NMN In In In OOOInCDOBInBOBOOLnOInOOON MMMNMMNHNMNNMNHNMNNN In In InInInIn InN InB NInNIn InN BBB In N IOBNInH BH BO InInBIOCD MMOInB O O O MMMMMd’ MMM¢ MMMMMMd’ MMM In InIn In 0 InInBInInInO Inc} 0 InBlnBN O O O B MMMMMd’ MMMd‘ d’ MMMMMd‘ Md‘ M In In In InIn In In InBOOI-nBONOOOInBBO InBOOB MMd‘ ¢ MMd’ Md’ d’ MMMMd’ Md’ MMM InIn In InInIn InIn NHNMMBHNCOOBMNHIn N H ONHHHIONO\OMNO\HOO\O BOON MMMMMMMN MMMN MMN MMN MN Lmomoovmmn B O\ O\d’IDBBIn B BO\O\®HO\NO® :roaoxmommmco N N N N MMMMN MMMN‘xN MMMN MN mm IOMNGNBHNInCnOB moo msmmmmmooamoxoxaam BIOLn MMMMMMMN MMN N MMN MMN N N ZSWEQZZZWWZZZZZZEZWZ H NN H HH N N H H44 HI FHH H P4 I1 61 .4N 0-; 04C!) HIDUL'DUJ U) HH zmmzzmmzmmmzmzzmzmmm P = practicum students Il I2 S + first-year interns second-year interns senior staff *4 haw— APPENDIX H AUDIBILITY us .np. APPENDIX H AUDIBILITY Earlier studies using the Counseling Center tape library indicated that there had been difficulty in hear- ing what was on the tapes. To deal with this problem, two methods were employed: (1) decisions were made for inaudible segments in making the master tape, including (a) go to next audible 3-minute segment, (b) move up one therapy session, (c) move down one therapy session, (d) leave present tape in, and (2) raters were instructed to make audibility ratings after each tape segment (where "1" means poor, "2" means "OK," and "3" means fine). The results show that 80% of the segments had audibility of at least "2," 89% had audibility of at least "1.75," and 95% had audibility of at least "1.5." 100 APPENDIX I CLIENT CODE NAME, AGE, VERBAL SCORE ON COLLEGE QUALIFYING TEST (CQT), TOTAL SCORE ON CQT _.;._..-1_..___— 4,~.~-_. APPENDIX I Table I-l CLIENT CODE NAME, AGE, VERBAL SCORE ON COLLEGE QUALIFYING TEST (CQT), TOTAL SCORE ON CQT Group 1 clients age _yg£bg1 total CQT Earnest 19 56 160 Margie 2O 61 156 Kate 21 55 118 Ben 20 54 129 Sam 19 5O 1#6 Marilyn 19 74 153 Date 19 26 ”70 Ross 21 55 121 Jean 20 69 158 Many (a) 21 73 156 Coféee 18 70 159 Tom 20 52 155 Diane 2O 47 100 Richard 18 59 16} Jerry -- __ __- Key 19 71 151 David 19 51 147 Ann 21 58 154 Mary (b) 19 48 92 Day -_ -- --- 101 102 APPENDIX I—-Continued Grou 2 clients a e verbal total C T 35352 - ----- ~* 2?“ 66’“” 167“” ‘9‘ Peggy 19 47 117 Barbara 19 74 186 Leon 24 42 ‘99 Rollo 2O 66 175 Abby 21 70 151 Edith 2O 65 145 Mr. E 19 54 152 Beth 18 71 150 Shirley 19 45 108 Elaine 21 40 141 Gary 19 57 156 Joan -- -- --— Henry 20 67 147 Jim 20 66 149 Sandy 19 70 180 Mr. C 18 65 155 Trudy 19 64 150 Wendy 20 60 118 Rita 18 62 152 Group 1 sum on CQT Verbal : 989; mean = 54.944 (n=18) Group 2 sum on CQT VBRbal = 1125;mean = 59.105 (n=19) t test on difference between means is non-significant. Group 1 sum on CQT Total = 2468gmean = 157.11 (n=18) Group 2 sum on CQT Total = 2856;mean = 144.0 (n=19) t test on difference between means is non-significant. Mean age of group 1 = 19.06 (n=18); mean age of group 2 clients = 19.68 (n=19). By inspection, there is no statistically significant difference. APPENDIX J GENERAL CHARACTERISTICS OF THE DATA, AND ADDITIONAL STATISTICALLY SIGNIFICANT CORRELATIONS APPENDIX J GENERAL CHARACTERISTICS OF THE DATA, AND ADDITIONAL STATISTICALLY SIGNIFICANT CORRELATIONS Grand means for the entire population are as follows: (a) (b) (C) (d) early session mean empathy = 3.12 late session mean empathy = 3.18 combined session mean empathy = 3.15 early session peak empathy = 3.67 late session peak empathy = 3.71 combined session peak empathy = 3.69 early session basal empathy = 2.52 1 late session basal empathy = 2.62 a combined session basal empathy = 2.57 mean empathy across time = +.0675 peak empathy across time = +.0437 basal empathy across time = +.1000 peak empathy and basal empathy are significantly correlated for early, late, and sessions combined (5 = .32, .34, and .34 respectively are all sig- nificant beyond the .05 level). That is, the higher the basal empathy, the higher the peak empathy. the early basal empathy score is significantly correlated with the combined session peak empathy score (£'= .34; p < .05). That is, knowing a thera- pist's lowest empathy score in the early session can aid in predicting his average peak empathy over early and late sessions combined. late peak empathy is significantly correlated with combined session basal empathy (£.= .37; p_< .01). late empathy scores are unrelated to early empathy scores for mean, peak, and basal empathy (£>= .13, .16, and -.06 respectively). 103 APPENDIX K MEAN, PEAK, AND BASAL EMPATHY OF PRACTICUM STUDENTS, FIRST-YEAR INTERNS, SECOND-YEAR INTERNS, AND SENIOR STAFF EACH CONSIDERED SEPARATELY (ANALYSIS III) APPENDIX K MEAN, PEAK, AND BASAL EMPATHY OF PRACTICUM STUDENTS, FIRST-YEAR INTERNS, SECOND-YEAR INTERNS, AND SENIOR STAFF EACH CONSIDERED SEPARATELY (ANALYSIS III) Results reported here are only those that involve therapist experience as a variable (see section on Training Implications for discussion): (a) Anxious—depressed clients as compared to other clients receive higher late session mean empathy from practicum students and second-year interns, less from first-year interns, and an equal amount from senior staff. The client discomfort main effect is not statistically sig- nificant however (see Table K-l and Figure K-l). Table K-l Analysis of Variance of Mean Accurate Empathy in the Late Session by Sex of Client, Sex of Therapist, Experience of Therapist, and Client Group Status Source of Variation SS DF MS F Signif. of F Main Effects Sexcl .007 l .007 .104 -- Sexth .119 1 .119 1.657 -- Expth .743 3 .248 3.454 .034 Grst .038 1 .038 .536 -- 2-way Interactions sexcl sexth .020 1 .020 .275 -- sexcl expth .099 3 .033 .460 -- sexcl grst .288 l .288 4.015 .055 sexth expth .361 3 .120 1.679 —— sexth grst .001 1 .001 .016 ~- expth grst 1.436 3 .479 6.671 .003 Residual 1.506 21 .072 Total 4.646 39 .119 |._n O h 105 3.7Jr Group 1 3.5-‘— Mean 3.3~- .__ 1:: empathy in 3.1-— GrouP late session 2.9~- 2.7~- ' I 1 v practicum first-year second-year senior students interns interns staff Figure K-l. Mean late session empathy broken down by therapist experience and client group status. (p_ < .01) (b) During the late session, senior staff and second-year interns achieve slightly higher basal scores, practicum students achieve much higher basal empathy, and first-year interns offer lower basal empathy with anxious-depressed clients as compared with clients low in subjective discom— fort (see Figure K—2). A similar pattern exists in sessions combined (E < .01). (c) Practicum students and senior staff offer equally high mean empathy in the late session, with interns less than the former two experience levels (p < .05). That is, late session mean empathy is high for practicum stu- dents, drops dramatically for first—year interns, then increases with experience until reaching a level similar to the starting point (see Figure K—3). 3.4~- 3.25- 3.0‘r Basal empathy 2'8 Igte 2.6~r Group 1 . Group 2 sess1on 2.4—_ 2.2‘* 2.0:5 1.8:: V 1 f practicum first-year second-year senior students interns interns staff Figure K-2. Basal late session empathy broken down by therapist experience and client group status (E.< .01) 3.4-~ Mean 3°3_h empathy 3 2-4 in ‘ late 3 11_ session ' 3.0—4 1 4 ~4 : practicum first-year second—year senior students interns interns staff Figure K-3. Late session mean empathy broken down by therapist experience. (p < .05) 107 (d) In combined sessions, practicum students and second— year interns avoid low basal empathy slightly better with female clients than with male clients, while first-year interns avoid low basal empathy better with male clients than with female clients. Senior staff seem to offer much lower basal empathy with male clients than with female clients (see Figure K—4). 2.8‘F 2°7~L female Basal clients empathy 2.6" in sessions 2.5._ combined 2.4-~ 2.34t male ients 2.2—4.— practicum first-year second-year senior students interns interns staff Figure K-4. Combined session basal empathy broken down by therapist experience and sex of client. (2 < .05) (e) The greater the experience level of the therapist, the higher the late session peak empathy (EF.334; p_< .05). (f) The greater the experience level of the therapist, the higher the late session mean empathy (£;.218; p_< .10). (g) There is no difference across time on peak empathy for practicum students with clients of either sex. Female first-year interns achieve higher peak empathy in late 108 therapy, while the male first-year interns offer higher peak empathy in early sessions. For second-year interns, male therapists offer higher peak empathy and female therapists lower peak empathy across time. Male senior staff achieve higher peak empathy during late sessions, while female senior staff offer the same peaks across time (see Figure K—5). +.44- +.3:- ale Peak 2__ therap1sts empathy . Female across «F . time +.l therapists 0 -------------- 0 change -.l~- _.2 T- -.3J- practicum first-year second-year senior students interns interns staff Figure K—5. Peak empathy across time broken down by sex of therapist and therapist experience. (p,< .05) (h) There is a strong statistical trend (p=.06) that is worth reporting: 1. Practicum students increase mean empathy across time with group 1 clients, and decrease mean empathy across time with group 2 clients. 2. First-year interns decrease mean empathy across time to group 1 clients, and have no change in mean empathy across time with group 2 clients. Mean empathy across time 109 Second-year interns, like practicum students, increase mean empathy across time with group 1 and decrease mean empathy across time with group 2. The extent of the change, however, is not as great as that made by practicum students. Senior staff increase mean empathy across time to both group 1 and 2 clients (see Figure K-6). Group 1 L I V r I I practicum first-year second-year senior students interns interns staff Figure K—6. Mean empathy across time broken down by therapist experience and client group status. (p_ < .10) APPENDIX L EMPATHIC UNDERSTANDING IN INTERPERSONAL PROCESSES APPENDIX L EMPATHIC UNDERSTANDING IN INTERPERSONAL PROCESSES. l A Scale for Measurement Robert R. Carkhuff Level I The verbal and behavioral expressions of the first person either do not attend to or detract significantly from the verbal and behavioral expressions of the second person(s) in that they communicate significantly less of the second person's feelings than the second person has communicated himself. The present scale "Empathic Understanding in In— terpersonal Processes" has been derived in part from "A Scale for the Measurement of Accurate Empathy" by C. B. Truax which has been validated in extensive process and outcome research on counseling and psychotherapy (summar— ized in Truax and Carkhuff, 1967) and in part from an earlier version which has been validated in extensive pro- cess and outcome research on counseling and psychotherapy (summarized in Carkhuff and Berenson, 1967). In addition, similar measures of similar constructs have received exten- sive support in the literature of counseling and therapy and education. The present scale was written to apply to all interpersonal processes and represent a systematic attempt to reduce the ambiguity and increase the relia- bility of the scale. In the process many important delin- eations and additions have been made, including in partic- ular the change to a systematic focus upon the additive, subtractive or interchangeable aspects of the levels of communication of understanding. For comparative purposes. Level 1 of the present scale is approximately equal to Stage 1 of the Truax scale. The remaining levels are approximately correspondent: Level 2 and Stages 2 and 3 110 111 Examples: The first person communicates no awareness of even the most obvious, expressed surface feel- ings of the second person. The first person may be bored or disinterested or simply operat- ing from a preconceived frame of reference which totally excludes that of the other person(s). In summary, the first person does everything but express that he is listening, understanding or being sensitive to even the feelings of the other person in such a way as to detraCt significantly from the communications of the second person. Level 2 While the first person responds to the expressed feelings of the second person(s), he does so in such a way that he subtracts noticeable affect from the communications of the communications of the second person. Examples: The first person may communicate some awareness of obvious surface feelings of the second per- son but his communications drain off a level of the affect and distort the level of meaning. The first person may communicate his own ideas of what may be going on but these are not con— gruent with the expressions of the second person. In summary, the first person tends to respond to other than what the second person is expressing or indicating. Level 3 The expressions of the first person in response to the ex- pressed feelings of the second person(s) are essentially interchangeable with those of the second person in that they express essentially the same effect and meaning. of the earlier version; Level 3 and Stages 4 and 5; Level 4 and Stages 6 and 7; Level 5 and Stages 8 and 9. The levels of the present scale are approximately equal to the levels of the earlier version of this scale. 112 Example: The first person responds with accurate under- standing of the surface feelings of the second person but may not respond to or may misinter— pret the deeper feelings. The summary, the first person is responding so as to neither subtract from nor add to the expressions of the second person; but he does not respond accurately to how that person really feels beneath the surface feelings. Level 3 constitutes the minimal level of facilitative interpersonal functioning. Level 4 The responses of the first person add noticeably to the expressions of the second person(s) in such a way as to express feelings of level deeper than the second person was able to express himself. Example: The facilitator communicates his understanding of the expressions of the second person at a level deeper than they were expressed, and thus enables the second person to experience and/or express feelings which he was unable to express previously. In summary, the facilitator's responses add deeper feeling and meaning to the expressions of the second person. Level 5 The first person's responses add significantly to the feel- ing and meaning of the expressions of the second person(s) in such a way as to (1) accurately express feelings levels below what the person himself was able to express or (2) in the event of ongoing deep self-exploration on the second person's part to be fully with him in his deepest moments. Examples: The facilitator responds with accuracy to all of the person's deeper as well as surface feel- ings. He is "together" with the second person or "tuned in" on his wave length. The facili- tator and the other person might proceed 113 together to explore previously unexplored areas of human existence. In summary, the facilitator is responding with a full awareness of who the other person is and a comprehensive and accurate empathic understanding of his most deep feelings. REFERENCES REFERENCES Abeles, N. The concept of therapeutic sensitivity and its relationship to training. Paper presented at the meeting of the American Psychological Association, Philadelphia, 1963. Alexander, J. F. Perspectives of psychotherapy process: Dependency, interpersonal relationships, and sex differences. Unpublished doctoral dissertation, Michigan State University, 1967. Anderson, S. Effects of confrontation by high- and low- functioning therapists. Journal of Counseling Apfelbaum, D. Dimensions of transference igtpsychotherapy. Berkeley: University of California Press, 1958. (Cited in Alexander, 1967.) Bare, L. Relationship of counselor personality-client personality similarity to selected counseling criteria. Journal of Counseling Psychology, 1967, 14, 419-425. Barron, F. An ego-strength scale which predicts response to psychotherapy. Journal of Counseling Psychology, 1953, 11, 327-333. Beck, A4 T. The diagnosis and management of depression. Philadelphia: University ofiPennsylvania Press, 1973. Becker, J. Depression: Theory and research. New York: John Wiley, 1974. Bergin, A. E. The evaluation of therapeutic outcomes. In A. E. Bergin and S. L. Garfield (Eds.), Hand- book of psychotherapy and behavior change: An empirical analysis. New York: John Wiley, 1971. 114 115 Bergin, A. E., & Garfield, S. L. (Eds.). Handbook of . psychotherapyand behavior change: An empirical analysis. New York: John Wiley, 1971. Bergin, A. E., & Jaspar, L. G. Correlates of empathy in psychotherapy: A replication. Journal of Abnormal and Social Psychology, 1969, 24, 477-481. Bergin, A. E., & Solomon, 8. Personality and performance correlates of empathic understanding in psycho- therapy. In J. T. Hart & T. M. Tomlinson (Eds.), New directions in client-centered therapy. Boston: Houghton-Mifflin, 1970. Bergin, A. E., & Suinn, R. Individual psychotherapy and behavior therapy. In M. Rosenzweig & L. Porter (Eds.), Annual review of psyghology (Vol. 26). Palo Alto: Annual Reviews Inc., 1975. Betz, B. Studies of the therapist's role in the treatment of the schizophrenic patient. American Journal of Psychiatry, 1967, 123, 963-971. Beutler, L., Johnson, D., Neville, C., & Workman, S. Accurate Empathy and the A-B dichotomy. Journal 9f Consulting and Clinical Psychology, 1972, 38, 372-375. Beutler, L., Johnson, D., Neville, C., & Workman, S. Some sources of variance in Accurate Empathy. Journal of Consulting and Clinical Psychology, 1973, 4Q, 167-169. Bienenfeld, S. Perspectives on psychotherapy process: Effects of consistency of therapeutic behaviEr and therapist characteristics. Unpublished masterTgithesis, Michigan State University, 1975. Black, J. D. Adjectives associated with various MMPI codes. In G. 8. Welsh & W. G. Dahlstrom (Eds.), Basic readings on the MMPI in psychology and medicine. Minneapolis: University of Minnesota Press, 1956. (a) Black, J. D. MMPI results for fifteen groups of female college students. In G. S. Welsh & W. G. Dahl- strom (Eds.), Basic readings on the MMPI in psy- chology and medicine. Minneapolis: UniVersity of Minnesota Press, 1956. (b) 116 Blumberg, R. Therapist leadership and client dogmatism in a therapy analogue. Psychotherapy: Theopy, Research, and Practice, 1972, 9, 132-134} Boerger, A., Graham, J., & Lilly, R. Behavioral correlates of single-scale MMPI code types. gpurnal of Con- sulting and Clinical Psychology, 1974, 42, 398-402. Bordin, E. Research strategies in psychotherapy. New York: John Wiley and Son, 1974. Bozarth, J. D., & Krauft, C. C. Accurate Empathy ratings: Some methodological considerations. Journal of Clinical Psychology, 1973, 28, 408-410. Campbell, J. H., & Rosenbaum, C. P. Placebo effect and symptom relief in psychotherapy. Archives of General Psychiatry, 1967, 16, 364-368. (Cited in Meltzoff & Kornreich, 1970.) Canter, A. The efficacy of a short form of the MMPI to evaluate depression and morale loss. Journal of Consulting and Clinical Psychology, 1960: 24, 14-17. Carkhuff, R. R. Helping and human relations (Vols. I & II). New York: Holt, Rinehart and Winston, 1969. Carkhuff, R., & Alexik, M. Effect of client depth of self-exploration upon high- and low-functionings counselors. Journal of Counseling Psychology, 1967, 14, 350-355. Carkhuff, R., & Berenson, B. Beyond counseling and therapy. New York: Holt, Rinehart, and Winston, 1967. Carkhuff, R., Kratochvil, D., & Friel, T. The effects of professional training: The communication and discrimination of facilitative conditions. Journal of Counseling Psychology, 1968, 15, 68-74. Cartwright, R. D., & Lerner, B. Empathy, need to change, and improvement with psychotherapy. Journal of Consulting PsychologY: 1963, 21, 138-144. Caudillo, Rev. C. A study of empathy in relation to personality traits measured by the MMPI with a population of graduate students in counseling psychology. Dissertation Abstracts International, 1972, 33 (lo-B), 4990. 117 Chinsky, J., & Rappaport, J. Brief critique of the mean- ing and reliability of Accurate Empathy ratings. Psychological Bulletin, 1970, 1;, 379-382. Cooke, J. MMPI in actuarial diagnosis of psychological disturbance in college males. Journal of Coun- seling Psychology, 1967, 14, 474-477. Costello, C. G. (Ed.). Symptoms of psychopathology. New York: John Wiley, 1970. Dahlstrom, W. G., & Welsh, G. S. An MMPI handbook: A gpide to use in clinical practice and research. Minneapolis: University of Minnesota Press, 1960. Dahlstrom, W. G., Welsh, G. S., & Dahlstrom, L. E. Ag MMPI handbook (Vol. 1). Minneapolis: University of Minnesota Press, 1972. Dietzel, C. Client-therapist complementarity and thera- peutic outcome. Unpublished doctoral dissertation, Michigan State University, 1971. Drake, L. E. Interpretation of MMPI profiles in counsel- ing male clients. Journal of Counseling Psychology, 1956, 3, 83-88. Drake, L. E. MMPI profiles and interview behavior. In G. S. Welsh & W. G. Dahlstrom (Eds.), Basic read- ings on the MMPI in psychology and medicine. Minneapolis: UniVersity of Minnesota Press, 1956. Drake, L. E., & Oetting, E. R. An MMPI codebook for counselors. Minneapolis: UniverSity of Minnesota Press, 1959. Ewing, T. Changes during counseling appropriate to the client's initial problem. Journal of Counseling Psychology, 1964, 11, 146-150. Fernbach, R. Authoritarianism: A selection variable for psychotherapy. Journal of Counseling Psychology, 1973' 22, 69-720 Fischer, J. Training for effective therapeutic practice. Psychotherapy: Theogy, Research, and Practice, 1975, 12, 118-124. ‘— Frank, J., Gliedman, L., Imber, S., Nash, E., & Stone, A. Why patients leave psychotherapy. Archives of Neurology and Psychiatry, 1957, 11, 283-299. 118 Fuller, F. F. Influence of sex of counselor and of client on client expressions of feeling. Journal of Counseling Psychology, 1963, $2! 34-40: Gallagher, J. MMPI changes concomitant with client- centered therapy. In G. S. Welsh & W. G. Dahl- strom (Eds.), Basic readings on the MMPI in_psy- chology and medicine. Minneapolis: University of Minnesota Press, 1956. Gilberstadt, H., & Duker, J. A handbook for clinical and actuarial MMPI interpretation. Philadelphia: W. B. Saunders, 1965. Gliedman, L., Nash, E., Imber, 8., Stone, A., & Frank, J. Reduction of symptoms by pharmacologically inert substances and by short-term psychotherapy. Archives of Neurology and Psychiatry, 1958, 19, 345-351. Gliedman, L. H., Stone, A. R., Frank, J. D., Nash, E. H., & Imber, S. D. Incentives for treatment related to remaining or improving in psychotherapy. American Journal of Psychotherapy, 1957, 11, 589-598. Goodstein, L. D. Regional differences in MMPI responses among male college students. In G. S. Welsh & W. G. Dahlstrom (Eds.), Basic readings on the MMPI in psychglogy and medicine. Minneapolis: Uni? versity of Minnesota Press, 1956. Gurman, A. Instability of therapeutic conditions in psy- chotherapy. Journal of Counseling Psychology, 1973' 29., 16-240 Guthrie, G. M. Common characteristics associated with frequent MMPI profile types. In G. S. Welsh & W. G. Dahlstrom (Eds.), Basic readings on the MMPI in psychology and medicine. Minneapolié: University of Minnesota Press, 1956. Gynther, M. D., Miller, F. T., & Davis, H. T. Relations between needs and behaviors as measured by the EPPS and ICL. Journal of Social Psychology, 1962, 51, 445-451. Hammer, E. F. (Ed.). Use of interpretation in treatment. New York: Grune and Stratton, 1968. 119 Hartley, R. E., & Allen, R. M. The MMPI and the EPPS: A factor analytic study. Journal of Social Psychology, 1962, 88, 153-162. Hartzell, J. P. A preliminary study of nurturant and/or aggressive therapists' responsiveness to expres- sions of dependency and hostility in the initial phase of psychotherapy. UnpubliShed doctoral dissertation, Michigan State University, 1967. Hathaway, 8. R., & McKinley, J. C. Scale 2 (Depression). In G. S. welsh & W. G. Dahlstrom (Eds.), Basic readings on the MMPI in psycholqu and medic1ne. Minneapolis: University of Minnesota Press, 1956. Hathaway, S. R., & Meehl, P. E. Psychiatric implications of code types. In G. S. Welsh & W. G. Dahlstrom (Eds.), Basic readings on the MMPI in psychology and medicine. Minneapolis: University of Minne- sota Press, 1956. Heath, H., & Korchin, S. Clinical judgments and self- ratings of traits and states. Archives of General Psychiatry, 1963, 8, 390-399. Heck, E., & Davis, C. Differential expression of empathy in a counseling analogue. Journal of Counseling Psychology: 1973, 88, 101-104. Hill, Clara E. Sex of client and sex and experience level of counselor. Journal of Counseling Psychology, 1975, 83, 5-11. Hogan, R. Development of an empathy scale. Journal of Consulting and Clinical Psychology: 1969, 88, Jones, L. K. Toward more adequate selection criteria: Correlates of empathy, genuineness, and respect. Counseling, Education and Supervision, 1974, is, 13—21. Karl, N., & Abeles, N. Psychotherapy process as a function of the time segment sampled. Journal of Consulting and Clinical Psychology, 1969, 133, 207-212. Kiesler, D. The process of psychotherapy. Chicago: Aldine Publishing Co., 1973. 120 Kleinmuntz, B. Annotated bibliography of MMPI research among college populations. Journal of Counseling Psychology: 1962, 8, 373-396. Kleinmuntz, B. Identification of maladjusted college stu- dents. Journal of Counseling Psychology, 1960, 8, 209-211. Kurtz, R. A comparison of different approaches to the measurement of counselor empathy_in personal counseling. Unpublished doctoral di§sertatiOn, Michigan State University, 1970. Kurtz, R., & Grummon, D. Different approaches to the measurement of therapist empathy and their relation to therapy outcomes. Journal of Con- sulting and Clinical Psychology, 1972, 88, 106-115. Langer, L. The implications for construct validisy and Rogerian Theory in three measures of empathy. Unpublished master‘s thesis, MiCHigan State University, 1972. Lauver, P., Kelley, J., & Froehle, T. Client reaction time and counselor verbal behavior in an interview setting. Journal of Counseling Psychology, 1971, 88, 26-30. 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. Levis, D. J., & Carrera, R. Effects of ten hours of implosive therapy in the treatment of outpatients. Journal of Abnormal Psychology, 1967, 18, 504-508. (Cited in Meltzoff & Kornreich, 1970.) Love, J. C. Research in psychotherapy: The initial interview as a microcosm of the total therapy sequence. Unpublished doctoral dissertation, Michigan State University, 1971. Luborsky, L., Chandler, M., Auerbach, A., & Bachrach, H. Factors influencing outcome of psychotherapy: A review of quantitative research. Psychological Bulletin, 1971, 18, 145-185. McNally, H. A., & Drummond, R. Ratings of Carkhuff's facilitative conditions: A second look. Counselor Education and Supervision, 1974, 18, 73-75 0 121 Marks, P., & Seeman, W. An atlas for use with the MMPI: Actuarial description of abnormal personality. Baltimore: Williams and Wilkins, 1963} Martin, P. L., & Toomey, T. C. Perceptual orientation and empathy. Journal of Consulting and Clinical Psy- chology, 1973, 88, 313. Matarazzo, R. G. Research on the teaching and learning of psychotherapeutic skills. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psychotherapy and behavior change: An empiriCal analysis. New York: John Wiley, 1971. Meikle, 8., & Gerritse, R. MMPI "cookbook" pattern fre- quency in a psychiatric unit. Journal of Clinical Psychology, 1970, 88, 82-84. Mello, N., & Guthrie, G. MMPI profiles and behavior in counseling. Journal of Counseling Psychology, 1958, 8, 125-129. Melloh, R. A. Accurate Empathy and counselor effective- ness. Dissertation Abstracts, 1965, 88, 7110. Meltzoff, J., & Kornreich, M. Research in psychotherapy. New York: Atherton Press, 1970. Merrill, R., & Heathers, L. The relation of the MMPI to the EPPS on a college counseling center sample. Journal of Consulting and Clinical Psychology, 1956, 88, 310-312. Mezzich, J., Damarin, F., & Erickson, J. Comparative validity of strategies and indices for differ- ential diagnosis of depressive states from other psychiatric conditions using the MMPI. Journal of Consulting and Clinical Psychology, 1974, 88, 691-698. Mitchell, R. Relation between therapist response to therapiSt—relevant client expressions and therapy process and client outcome. Unpublished doctoral dissertation, Michigan State University, 1971. 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. 122 Mullen, J., & Abeles, N. Relation of liking, empathy, and therapist's experience to outcome of therapy. Journal of Counseling Psychology, 1971, 88, 39-43. Page, H. A. An assessment of the predictive value of certain language measures in psychotherapeutic counseling. In W. U. Snyder (Ed.), Group report of a program of research in psychotherapy (Mimeo- graphed). University Park, Pa.: Pennsylvania State College, 1953, 88-93. (Cited in Meltzoff & Kornreich, 1970.) Patterson, C. H. A current View of client-centered or relationship therapy. The Counseling Psycholggist, 1969, 8, 2—25. Persons, R., Persons, M., & Newmark, I. Perceived helpful therapist's characteristics, client improvement, and sex of therapist and client. Psychotherapy: Theory, Research and Practice, 1974, 88, 63-65. Pierce, R., Carkhuff, R., & Berenson, B. The differential effects of high and low functioning counselors upon counselors-in-training. Journal of Clinical Psychology, 1967, 88, 212-215. Plyler, S. A. Personality differences as measured by the MMPI between diagnosed groups of male clients seen at a college counseling bureau. Unpublished doc- toral dissertation, University of Missouri, 1965. Rappaport, J., & Chinsky, J. Accurate Empathy: Confusion of a construct. Psychological Bulletin, 1972, 88, 400-404. Riess, B. F. Some causes and correlates of psychotherapy termination: A study of 500 cases. International Mental Health Research Newsletter, 1972, 88, 4-7. (Cited in Bergin & Suinn, 1975.) Rogers, C. R. The necessary and sufficient conditions of therapeutic change. Journal of Consulting Psy- chology, 1957, 88, 95-103. Rogers, C. R. (Ed.). The therapeutic relationsh8p with schizophrenics. Madison, Wisc.: The University of Wisconsin Press, 1967. Rogers, C. Empathic: An unappreciated way of being. The Counseling Psycholggist, 1975, 8, 2-10. 123 Rosenzweig, M., & Porter, L. (Eds.). Annual review of psychology, 88. Palo Alto, Calif.: Annual Reviews, Inc., 1975. Rosenzweig, S., & Folman, R. Patient and therapist variables affecting premature termination in group psychotherapy. Psychotherapy, 1974, 88, 76-79 0 Safran, J. A comparison of the personal preferences and self concept of empathic and non-empathic coun- selor education students. Dissertation Abstracts International, 1973, 88 (ll-A), 6099-6100. Schauble, P., & Pierce, R. Client in-therapy behavior: A therapist guide to progress. Psychotherapy, 1974, 88, 229-234. Scher, M. Verbal activity, sex, counselor experience, and success in counseling. Journal of Counseling Psychology, 1975, 88, 97-101. Scott, R. W., & Kemp, D. E. A-B Scale and empathy, warmth, genuineness, and depth of self-exploration. Journal of Abnormal and Social Psycholegy, 1971, 88, 49-51. Shapiro, J. G., Foster, C., & Powell, T. Facial and bodily cues of genuineness, empathy, and warmth. Journal of Clinical Psychology, 1968, 88, 233-236. (a) Shapiro, J. G. Relationship between visual and auditory cues of therapeutic effectiveness. Journal of Clinical Psychology, 1968, 88, 236-239. (b) Snett, P. The effect of therapistepatient conflict simi— larity upon therspiSts1 prognostic evaluations and other clinical judgments of their_patients. Unpublished doctoral dissertation, Michigan State University, 1972. Snyder, W. U. The psychotherapy relationship. New York: The MacMillan Company, 1961. Stein, L., Green, B., & Stone, W. Therapist attitudes as influenced by A-B therapist type, patient diagno— sis, and social class. Journal of Consulting and Clinical Psychology, 1972, 88, 301—307. 124 Stone, A., Frank, J., Nash, E., & Imber, S. An intensive five-year follow-up study of treated psychiatric patients. Journal of Nervous and Mental Disease, 1961, 133, 410-422. Strupp, H., & Bloxom, A. An approach to defining a patient population in psychotherapy research. Journal of Counseling Psychology, 1975, 88, 231-237. Strupp, H., Wallach, M. S., Jenkins, J. W., & Wogan, M. Psychotherapists' assessments of former patients. Journal of Nervous and Mental Disease, 1963, 881, 222-230. (Cited in Bordin, 1974.) Taulbee, E. S. Certain personality variables and continu- ation in psychotherapy. Journal of Consulting and Clinical Psychology, 1958, 88, 83-89. Truax, C. B. Effective ingredients in psychotherapy: An approach to unraveling the patient-therapist interaction. Journal of Counseling Psychology, 1963, 88, 256-263. Truax, C. B. Influence of patient statements on judgments of therapist statements during psychotherapy. Journal of Clinical Psychology, 1966, 88, 335-336. Truax, C. B. The meaning and reliability of Accurate Empathy Ratings: A rejoinder. Psychological Truax, C. B., & Carkhuff, R. R. Toward effective counsel- ing and psychotherapy. Chicago: Aldine Publish— ing, 1967. Truax, C. B., & Wargo, D. G. Psychotherapeutic encounters that change behavior: For better or worse. American Journal of Psychotherapy, 1966, 88, 499-522. Truax, C. B., Wargo, D., Frank, J., Imber, S., Battle, C., Hoehn-Saric, R., Nash, E., & Stone, A. The therapist's contribution to Accurate Empathy, Nonpossessive Warmth, and Genuineness in psycho- therapy. Journal of Clinical Psychology, 1966, 88, 331-334. Truax, C. B., & Wittmer, J. The effects of therapist focus on patient anxiety source and the interaction with therapist level of Accurate Empathy. Journal of Clinical Psychology, 1971, 81, 297-299. 125 Van der Veen, F. Effects of the therapist and the patient on each other's behavior. Journal of Consulting and Clinical Psychology, 1965, 88, 19-26. Venema, J. The effects of expectancy training, commitment, and therapeutic conditions upon attrition from psychotherapy. Paper presented to the American Psychological Association, Miami, Florida, September, 1970. Vesprani, G. Accurate Empathy in a college companion pro- gram. Dissertation Abstracts, 1968, 887B, 3483. Watts, G. P. The Carkhuff Discrimination Scale as a pre- dictor of Accurate Empathy. Journal of Consulting and Clinical Psychology, 1973, 88, 202-206. Wechsler, H., Grosser, G., & Busfield, B. The Depression Rating Scale. Archives of General Psychiatry, 1963, 8, 334-343. Welsh, G. S., & Dahlstrom, W. G. (Eds.). Basic readings on the MMPI in psychology and medicine. Minneapo- lis: University of Minnesota Press, 1956. Zimmerman, F. The relationship between counselor accuracy in perceiving affect and client perceived empathy. Dissertation Abstracts International, 1973, 88 (8-AY, 4109. Zubin, J., & Fleiss, J. Current biometric approaches to depression. In R. R. Fieve (Ed.), Depression in the 1970's, Excerpta Medica, 1971. Zubin, J., Salzinger, K., Fleiss, J., Gurland, B., Spitzer, R., Endicott, J., & Sutton, S. Biometric approach to psychopathology. In M. Rosenzweig & L. Porter (Eds.), Annual Review of Psychology, 1975, 88, 629-636. Zucker, R. A., & Manosevitz, M. MMPI patterns of overt male homosexuals: Reinterpretation and comment on Dean and Richardson's study. Journal of Consulting and Clinical Psychology, 1966, 88, 555-557. Zuckerman, M., & Monashkin, I. Self-acceptance and psycho- pathology. Journal of Consulting and Clinical Psychology, 1957, 88, 145—148. IIIJIIIIHIHHIIWIWHIHHlHlleJJHIIHIJIIIIHJIIHWI