1111111111111 \1111‘1\1\1‘211\1\\11\\\\ ~ 3 1293 10591 9934 ‘IV1ESI.J RETURNING MATERIALS: P1ace in book drop to LJBRAfiJES remove this checkout from JI-IKSIIIL your record. FINES wi11 be charged if book is returned after the date stamped be1ow. AN ANALYSIS OF THERAPIST AND CLIENT VERBAL RESPONSES IN SUCCESSFUL AND UNSUCCESSFUL INSIGHT-ORIENTED PSYCHOTHERAPY By Timothy T. Eaton A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1984 f (.f‘ L [vi '1) Abstract AN ANALYSIS OF THERAPIST AND CLIENT VERBAL RESPONSES IN SUCCESSFUL AND UNSUCCESSFUL INSIGHT-ORIENTED PSYCHOTHERAPY By Timothy T. Eaton The primary purpose of this study was to determine whether the frequency of verbal response mode category use by therapists and clients can be utilized to differentiate successful and unsuccessful insight-oriented psychotherapy. A second feature of the study focused on identifying thera- ‘pist response ‘modes that are most effective in eliciting individual client response modeso Rather than correlating therapist and client verbal responses with measures of therapeutic success as has been attempted in the past, this study was designed to directly compare response mode fre- quencies between four groups of cases developed around successful-unsuccessful and nonpathological-pathological dimensions. A step-wise multiple regression analysis was also employed to identify the relationships between therapist and client response modes within each group. The principle hypotheses were stated as follows: 1) therapist response modes would not be useful in differentiating levels of success; 2) client response modes defined as uncovering responses would be found in greater frequency in successful cases; and 3) the therapist response modes of interpretation and reflection would be most positively associated with client uncovering responses. The first hypothesis was supported, but the second and third hypotheses were not. Client uncover- ing responses were not found in greater frequency in success— ful cases, and no clear relationships were found between therapist response modes and client uncovering responses. These results more clearly define the role of verbal response mode category systems as descriptive measures of therapy process, and suggest the importance of the therapist's role in psychotherapy as more than simply a facilitator of emo- tional arousal and client experiencing. To Kay and to my parents iii ACKNOWLEDGEMENTS I would like to express my deepest gratitude to my thesis committee. Dr. Norman Abeles, Dr. Bertram Karon, and Dr. G. Anne Bogat have displayed sincere interest in psychotherapy process research that is not often found even among clinical psychology faculty. Dr. Abeles, Chairman of my thesis committe, has given me the personal support, acceptance, and respect needed to make this research possible. His experience and expertise in psychotherapy process research has contributed greatly to the development and completion of the study, and his concern with the continued evaluation and improvement of psycho- therapy has been especially motivating. Dr. Karon's wit, vitality, and renowned expertise as a psychotherapist have made a significant impact on my development as a clinical psychologist, and on my attitudes and outlook toward the process of psychotherapy. Dr. Bogat's concern and genuine care for the education and training of clinical psychologists has had great impact on me, and her emphasis on the production of quality research and thought has had a great influence on the development and completion of this work. I want to especially thank my research assistants and iv raters, John Batdorf, Matthew Daly, Susan Harris, Kirk Hubbard, Al Ko, Latonya Roseboro, Brad Thomasma, and Ann Wagner, for the interest and effort they put into this project. They were a pleasure to work with, and should be a credit to any profession they pursue I would like to thank James Thomas, a good friend and astute scholar, for his timely advice and unselfish wisdom. I would like to extend special thanks to my typist, Bea Archambeault, for her caring attitude and marvelous skills at completing the final draft of this thesis. Her concern with this piece of my life was far more than expected, and greatly appreciated. I would also like to thank my family, Mom, Dad, Stacie, and my grandparents, for the loving support and trust they have shown me. They allowed my goals to be within my grasp, and my achievements to feel especially meaningful. Finally, my deepest appreciation goes to my wife, Kay. Without her understanding, support, and love this work would not have been possible. My personal accomplishments truly seem unimportant compared to the life that we have together. TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . INTRODUCTION 0 O O O O O O O O O O 0 .V O O O 0 REVIEW OF THE LITERATURE . . . . . . . . . . Process Variables: A Brief Review. . . . Conclusion. . . . . . . . . . . . . . . . Verbal Response Modes: A Brief Review. HYPOTHESES . . . . . . . . . . . . . . METHOD . Data. . . . . . . . . . . . . . . . . Clients . . . . . . . Therapists. Instruments A. Measure of Pathology . . . . . . . B. Outcome Measures . . . . . . . . . C. Process Measures . . . . . . . Rating Groups . Data Sampling Procedures. General Procedure . RESULTS . . . . . . . . . Preliminary Comments. . . . . . . . . A. Calculation of Frequencies . .'. Expression of Reliability Coefficients B. C. Comparison of Groups and Tests of Significance .'. . D Analysis of the effects of mixed Outcome Cases. . . . . . . . . . . Part I: Comparison of Therapist and Client Verbal Response Mode Group Frequencies. . vi Page ix 48 49 49 53 Page Hypothesis I . . . . . . . . . . . . . . . . 53 Hypothesis II . . . . . . . . . . . . . . . . 56 Hypothesis III. . . . . . . . . . . . . . . . 59 Part II Multiple Regression with Therapist and Client Response Modes . . . . . . . . . . . . . . 62 Data Analysis . . . . . . . . . . . . . . . . 62 Hypothesis IV . . . . . . . . . . . . . . . . 69 Hypothesis V. . . . . . . . . . . . . . . . . 69 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . 71 I. Is the Quantification of Therapist Response Modes useful in Differentiating Successful and Unsuccessful Psychotherapy? . . . . . . . 72 II. Is the Quantification of Client Response Modes useful in Differentiating Successful and Unsuccessful Psychotherapy? . . . . . . . 74 III. Can Individual Therapist Response Modes be used to predict and elicit particular Client Response Modes? . . . . . . . . . . . . . . . 84 IV. What are the Implications of this Study for the use of Verbal Response Mode Category Systems?. . . . . . . . . . . . . . . . . . . 87 V. What are the Implications for future Research? . . . . . . . . . . . . . . . . . . 90 SUMMARY. . . . . . . . . . . . . . . . . . . . . . . . 94 REFERENCE NOTES. . . . . . . . . . . . . . . . . . . . 97 LIST OF REFERENCES . . . . . . . . . . . . . . . . . . 99 APPENDICES . . . . . . . . . . . . . . . . . . . . . . A. Therapist Response Mode Rating System . . . . 108 B. Client Verbal Response Category System. . . . 142 C. Sympton Distress Checklist (SCL-9OR) Client Form . . . . . . . . . . . . 151 vii Page Symptom Distress Checklist (SCL-9OA) Clinician Form . . . . . . . . . . 159 Posttherapy Client Questionnaire . . . . . . 161 Posttherapy Therapist Questionnaire. . . . . 168 Standard (T- Score) Norms for Michigan State Psychological Clinic Clinician Ratings of Male and Female Outpatients on the SCL- 90R' 5 Symptom Dimensions and Global Pathology Index. . . . . . . ... . . . . . . . . . . 174 viii 10. ll. 12. 13. 14. LIST OF TABLES Page Comprehensive process analysis model . . . . . . 14 An example of helper response mode systems . . . 16 Helpful and nonhelpful event clusters. . . . . . 22 Graphic display of helpful and nonhelpful event clusters . . . . . . . . . . . 26 Description of groups: Sex, mean number of sessions, and average pathology ratings. . . . . 39 Comparison of mixed cases with unsuccessful cases within Groups II and IV. . . . . . . . . . 51 Comparison of successful and unsuccessful cases across nonpathological and pathological groups . 54 Comparison between successful-nonpathlogical cases and consensus unsuccessful-nonpatholog- ical cases 9 c t. c o o o o o o o o o o O o o o 0 56 Comparison of successful and unsuccessful cases across nonpathological and pathological groups . 57 Comparison of successful and unsuccessful cases with mixed outcome cases removed . . . . . . . . 58 Comparison of successful and unsuccessful cases between nonpathological and pathological groups. 60 Individual case frequencies of client response modes in Group III . . . . . . . . . . . . . . . 61 Percentages of form and intent codes for three types of therapy . . . . . . . . . . . . . . . . 75 A description of consensus unsuccessful and mixed outcome cases of Group IV. . . . . . . . . 82 ix Page 15. Sequential analysis of the immediate effects of the counselor predominant response on the client's subsequent two response units. . . 85 INTRODUCTION Understanding the process of psychotherapy in terms of the "ingredients" needed for behavior change has been a major theme of research in psychotherapy for many years. In their introduction to their review of psychotherapy process variables as related to therapeutic outcome, Orlinsky and Howard (1978) aptly offer a brief commentary on the state of psychotherapy research: The extraordinary diversity that we find in psychotherapy research parallels (and to a large extent is the product of) the amazing variety of conceptualizations and procedures that define the clinical practice of psycho- therapy. There are, to name a few: psycho— analytic and neo-analytic therapies; behav- iorist and neo-behaviorist therapies; cogni- tive, emotive, and body therapies; verbal, activity, and play therapies; as well as combinations, permutations, eclectic inte- grations, and idiosyncratic syntheses. Among them all, there is no standard defin- ition of what occurs in, or is distinctive of, therapeutic rocess; no consensus about the intended effects of therapy, or the criteria of therapeutic outcome- hence, no agreement concerning the seIection and measure- ment of meaningful process and outcome variables. (p. 284) Even though each approach to psychotherapy can be characterized according to the particular "effective processes" or "change mechanisms" with which it is asso- ciated (Elliott et a1., 1982a), and every practicing or research psychotherapist must be concerned with these 1 particular mechanisms, little definitive research has resulted from such concern. Fault for the somewhat confused and convoluted field of psychotherapy process research cannot be placed on past and present researchers. By its very nature, process research is tedious and difficult, and possesses little attractive qualities for many budding psychologists. If fault must be found, the immaturity of the field must be largely to blame. Orlinsky and Howard (1978) describe the present position of psychotherapy process research as pre-paradigmatic. Elliott at al.4(l982a) view the field as being too confined to early models of research while ignoring more current developments in more recent research models. Whatever terminology one wishes to employ in describing the developmental position of psycho— therapy process research, the main conclusion is always that onew and innovative research is needed to not only add to the general body of knowledge that has thus far been developed, but to locate directions of importance for future research. REVIEW OF THE LITERATURE Process Variables: A Brief Review Even though psychotherapy process research is still in its infancy, a number of process variables have been explored providing a firm foundation from which to design further research and extend our base of knowledge. Orlinsky and Howard (1978) provide a comprehensive review of process variables associated with therapeutic outcome. Their review highlights the process variables concentrated upon thus far in psychotherapy research, and a summary of this research will be helpful as a base for future discussion. Many of the process variables thus far discovered and explored in psychotherapy research have been a direct result of the facilitative conditions espoused by Rogers (1957). Studies focusing on therapist empathy, uncondi- tional positive regard, and self-congruence all come from the perspective of the therapist's interpersonal behavior having a direct impact on behavior change. A number of measures have been developed in the attempt to define and understand these facilitative conditions, one of the most influential being the development of the rating scales by Truax and his colleagues (Truax and Carkhuff, 1967). Ratings for these scales are obtained through the use of nonparticipant observers rating recorded process segments 3 of therapy, offering a research technique that has been adapted for use in a number of various process measures. A great deal of the research has been focused upon the relationship between therapist warmth, empathy, and therapeutic outcome. A study by Halkides (1958) showing this to be a positive relationship has seemingly stirred a wealth of research studies. Truax and Mitchell (1971) discovered eleven studies on warmth and thirteen on accurate empathy in their review of the literature prior to 1970. Orlinsky and Howard's (1978) review added another twelve studies on warmth (e.g., Truax, 1970a; Mullen and Abeles, 1971; Truax, Wittmer, and Wargo, 1971; Garfield and Bergin, 1971; Schauble and Pierce, 1974; Truax et a1., 1973) and twenty-two on empathy (e.g., Ber- gin and Jasper, 1969; Truax, 1970b; Mintz, Luborsky, and Auerbach, 1971; Kurtz and Grummon, 1972; including the preceding studies on warmth). As a whole, the studies cited here and the overall review by Orlinsky and Howard lead to the conclusion that empathy and warmth do not necessarily lead to a positive outcome, but they do significantly add to the combination of significant "ingredients" in therapy that as a whole lead to a positive outcome. In another study of therapists' interpersonal behav- ior, Crowder (1972) used Leary's (1957) model to describe the relationship between interpersonal process and outcome in successful and unsuccessful cases. Crowder discovered that therapists in both successful and unsuccessful cases were most frequently supportive-interpretive in their interpersonal behavior, but therapists in the successful cases were significantly more supportive-interpretive, less hostile-competitive, and less passive-resistant late in therapy; and more hostile-competitive and less passive- resistant early in therapy than the therapists in the unsuccessful cases. Agreeing with Orlinsky and Howard's (1978) assessment of these findings, active and positive participation by the therapist seems to be of positive therapeutic value. In support of this general conclusion, a number of studies have linked active and positive participation by the therapist to successful outcome by studying the specific clinical tedhniques used by therapists. Ashby et a1. (1957) found leading or guiding behavior to be more associated with successful therapeutic outcome as compared with reflective behavior. Direct approval has been linked to successful outcome as opposed to clarify- ing or interpretive statements (Sloane et a1., 1975). Likewise, confrontation has been associated with positive therapeutic outcome (Mainard, Burk, and Collins, 1965). In opposition to some of these findings, Baker's (1960) study of leading and reflective techniques found no significant differences between the two, and Nagy (1973) found confrontation to be unrelated to therapeutic outcome. The amount and style of therapists' verbal activity have also been looked at as possible sources of positive therapeutic outcome. However, studies measuring the amount and rate of therapist speech have shown no relationship to therapeutic outcome (Barrington, 1961; Scher, 1975; Sloane et a1., 1975). On the other hand, research by Rice (1965) on therapists' voice quality has indicated that an ex- pressive vocal style and use of fresh language are associated with greater positive outcome as opposed to an artificial vocal style and use of stereotypic language. Systematic case-study research by Strupp (1980a, 1980b, 1980c) suggested that although therapists' skills and attitudes toward the patient do have some effect on the process of psychotherapy, the patient variables are really the key to therapeutic process associated with outcome. These results highlight the growing interest in client process variables as opposed to therapist variables in the attempt to understand the effectivenss of psycho- therapy. Studying client variables in process research follows the same investigative paths as studying therapist variables. For example, voice quality of clients who were successful in therapy has been studied showing that such clients use more of an open and expressive vocal style than less successful clients (Butler, Rice, and Wagstaff, 1962). Barrington's (1961) research suggests that clients who use more words with larger numbers of syllables tend to have more positive outcome. Furthermore, Orlinsky and Howard (1978) report a number of articles suggesting that successful clients not only have more to say in therapy, but also take the time to think about what they want to say (use of silences). Crowder's (1972) study of interpersonal behavior in psychotherapy reported earlier for therapist behaviors also found interesting client behaviors associated with more successful outcome. In the early phases of therapy, the more successful clients were more hostile-competitive, less passive-resistant, and more support-seeking. During the middle phases of therapy, the trends on the passive- resistant and support-seeking scales continued. Studies focusing on client self-perceptions and self- experience have provided some interesting data beyond the "objective" observer perspective. Lorr’and McNair (1964) found that clients who perceive themselves as acting in a hostile—controlling manner had less successful outcomes than those not having such self-perceptions. In the same study, they also found that clients who perceive themselves as being actively involved in therapy are more successful. Supporting this finding, Comes-Schwartz (1978) also found that greater patient involvement most consistently predict- ed positive therapeutic outcome. Saltzman et a1. (1976) found that clients who felt a greater sense of responsibility for solving their problems and changing their behavior offered higher self-rated out- comes. From a greater self-experiencing perspective, Cabral et a1. (1975) found that clients in group therapy who perceive themselves as having intense emotional express- ions were consistently more successful in their therapy. Client self-experiencing fits the theoretical model of good therapy process espoused by Eugene Gendlin (1973), and stimulated by the client-centered approach developed by Carl Rogers (1957). Gendlin dismisses the importance of any individual therapeutic perspective by concentrating on how therapy, no matter what the orientation, elicits an "experiencing” response from the client. This response, which may best be described as an accurate and personal feeling about some event, situation, or thought expressed in cognitive and affective terms (Rice, 1974), is the key to successful therapeutic outcome. Rice (1974) explains the "experiencing" response as a cognitive necessity to completely processing feelings that had been denied or distorted (thus incompletely processed) when first en- countered. However the ”experiencing" response is described, its importance as an insight event in psychotherapy has been verified. Gendlin et al. (1960) found that clients who move from talking about their feelings to experiencing them are more likely to improve within individual client- centered therapy, and no correlation was found between positive outcome and therapists' perceptions of clients talking about therapy, the therapist, or the present. Kirtner and Cartwright (1958) differentiated successful and unsuccessful cases of client-centered therapy by identifying those clients who discussed their feelings in the first session as opposed to those clients who spoke of their problems as being basically external to themselves. Truax and Wittmer (1971) obtained similar results when they found clients' use of personal references to be correlated with positive outcome as opposed to clients' use of nonpersonal references. These data are further verified by a study by Schauble and Pierce (1974) which identified an association between positive outcome and clients directly confronting their problems and feelings. A more recent study by Elliott et al. (19823) using four different evaluative paradigms of therapy process (process- outcome, sequential process, immediate process recall, and retrospective attribution) identified client experiencing and therapist's direct reference to that experiencing process as the primary helpful factors in a single case study. A number of studies have used more of a quantitative 10 approach in their study of client experiencing based on the scales developed by Gendlin et a1. (1968). Studies by Tomlinson and Stoler (1967); Tomlinson and Hart (1962); van der Veen (1967); Kirtner et a1. (1961); and Gendlin et a1. (1968) all relate successful outcome in psycho— therapy to client personal referents, internal referenc- ing, or what may generally be called the process of client experiencing. CONCLUSION: Generally, good therapy process can be described based on the types of process-outcome research studies that have been briefly reviewed to this point. Even though process research continues to struggle to define valid guidelines for effective psychotherapy, general trends have been established. Effective psychotherapy may best be distinguished by the collaborative bond built between client and thera- pist who both invest a great deal of effort and energy in making the relationship supportive and encouraging, but also challenging and stimulating (Orlinsky and Howard, 1978). The techniques the therapist uses are vital in promoting this positive relationship. In this sense, the orientation from which the therapist chooses to operate must largely be dependent on the therapist's personal feelings of what will be effective and comfortable for ll him or her. The lack of evidence for particular orienta- tions being more effective than others has been well- established (Smith, Glass, and Miller, 1980). Particular therapist techniques may also be used in the safe and supportive environment of the therapeutic alliance to stimulate the client to express and experience painful or frightening thoughts or feelings. Depending on the therapist's orientation, a variety of terms may be used to describe this activity, including: growth facili- tating, positive transference, working through corrective emotional experience, reciprocal inhibition, modeling, or positive reinforcement (Orlinsky and Howard, 1978). It is the study of this therapeutic activity that may offer the greatest challenges to current and future process research. Verbal Response Modes: A Brief Review Consider the following exchange between a therapist and client during an initial session: C: I just feel so low. It's very frightening, like the whole world is pressing in on me, and I don't know why I feel this way. T: You feel frightened and depressed, like the whole world is closing in around you, but you just can't pinpoint the reason why you feel this way. The therapist chose to reflect the client's 12 disclosure at this point even though a number of response options were open to him. He could have begun asking questions about the client's current situation, offered an interpretation concerning why the client feels this way, or he could have offered some advice for the client to follow outside the session to help alleviate the client's feelings. Which response would be best in this situation is one of the questions that can be addressed using the techniques of verbal response mode research. As just stated, the preceding question is one of many that can be asked about communication in a help- intended situation. Descriptive questions may be asked, such as: How do people communicate with each other in dyadic or small group situations?; What are the options or choices people have available to them in speaking?; What are the processes or rules people use to communicate to each other? Equally, prescriptive questions may be asked about help-intended communication, such as: How do people do things with words?; How do people accomplish psychological help using words?; What kinds of communica- tion are most helpful and under What circumstances?; How do we know whether a response is helpful or not?; What kinds of communication are psychologically harmful and in what circumstances?; What are the best methods to teach people to be psychological helpers? As illustrated above, verbal responses can be used 13 to describe the process of help-intended communication, whether it be psychotherapy or a conversation between friends, or they can be considered as a process variable in and of themselves. Russell and Stiles (1979) note three basic aspects of language in psychotherapy: content, action, and style. Elliott (Note 2) adds two or more, quality and state, to form a comprehensive model to describe the process of help-intended communication (see Table 1). As part of this model, one of the choices a therapist or helper makes is what they intend to do by what they say, or in other words, what action will their language take (Russell and Stiles, 1979). This choice of mode of action is referred to as a response mode, and is probably the most salient for therapists of the various aspects of helping process (Elliott et al., 1981b). Research using verbal response modes has a rela- tively limited history, but does extend back to the begin- nings of psychotherapy process research. Much of this research was descriptive in nature, and verbal response modes provided a system through which verbal communication could be classified. Bales (1950) used response modes in this manner as part of his process analysis of small group interactions. Strupp (1955) compared response mode use patterns of therapists from the client-centered school with therapists from the psychoanalytic tradition. Snyder 4 1... mammH wagowuo mmocaamaamm Hmnno>co xmwh mafia menace: coaumamm umfiamuon opoz omconmom \aowuo< knummEm nooomm Homaom wcfiaoom coauom uo>uomno mammH wcwowuo} wcwxuoz Hmnuo>coz xmms mafia mchxcass coaumamm ocmfiu 1 £55 H >833. wcwocowuomxm nomomm ncowummum>cou waHoo nowhmm o>wuommmuom ououm muwamsa mamum cowuo< .uaouaoo Homo: mwmhamz< mmmoOHm o>fimno£mhmfioo .H oHan umfimmuo£9 bcmaao 15 (1945) used response modes in a similar way when he investigated use patterns in nondirective psychotherapy. Using this early research as a beginning foundation, response modes began to be organized around systematic frameworks for use in training packages (Elliott et al., 1981b). These frameworks had limited utility, and re- search usingtflmmIfocused on validating the training package rather than attempting to understand the nature of the response modes. However, it was at this point that Goodman and Dooley (1976) developed their framework of verbal response modes (see Table 2) that attempted to in- tegrate the process research tradition with the more cur— rent training packages (Elliott et a1., 1981b). Their verbal response mode system could easily be used for research purposes, and offered a standard system that could be applied to any therapy or helping situation. Furthermore, a training package was developed with a con- cern for understanding the response modes and using them in the most helpful manner (Goodman, 1979). It was Goodman and Dooley's framework that seemed to spark the development of a host of verbal response mode classification systems for use in psychotherapy process research. Three popular systems include: Stiles' Eggbal Response Mode System (1978; 1979), Hill et al.'s Counselor and Client Verbal Response Category System (Hill, 1978; Hill et a1, 1981), and Elliott et al.'s Therapist Response Mode Rating System (Elliott, 1979a; Elliott et a1., 1982b). l6 pmamwooam uoz ucowau wCHEuowcH “amuse wcawcmaamso ucofiau osu wCHDHOQQDm maom mo wcfiumnm summonwaoa mamm posuo coHumEMomcH ucoEooummmHo mocmnsmmmom QHDmOHUwHQIMHmm wcwp3ouo .m> wcHBOHH<.HmnHm> cowuasnuoucH .m> oocoawm cu ucowau unachaaxm cowumuowduoucH mamm mo mewwmnm oumnonwaon whamoaomwoumaom wcwpcmum . lumps: waumoHcDEEoo coauooamom maom cu uCowHo wchHwHaxm cowumuoudnmucH cowmmom cwnuwz wcapflsu uCoEomw>p< mmoooum mewpcmum Cowmmom mo use wcflpwnu ucoEomH>p< Hmuocou upopc: wcwumowcsesoo cowuooamom cowumEpomcH wcwwosumo cowumoao pomoHo ucowau wawpwso uCoEomH>p< cowumenomcH wcfluofiumo coflumoso Como cofiumauomcH mnemonuwo cowumoso coaucoucH wcwmaom Nwowoumu cowucousH wawmamm Nuowoumo Amwaav smumsm ..Hm um buoaaam Aommav Eoumzm m.moHoonH new cwenoow mEouwmm opoz mmcoamom Hmaaom mo oHQmem c< .N oH£MH 17 These systems can all be used to categorize thera- pists' verbal responses in psychotherapy, and Stiles' and Hill's systems can be used to categorize clients' verbal responses in psychotherapy (Elliott has a modified version of Hill's client rating system, Note 3). All of the systems employ trained raters to do the actual response ratings, and all are flexible enough to be used in a variety of research endeavors. The unit of analysis used with the systems has generally been the verbal sentence or clause, but recent research comparing three unit types (clauses, sentences, and speaking turns) across the pre- ceding three category systems showed that unit type makes little difference to reliability or validity; however, response mode levels do change depending on the unit type used (Eaton et a1., Note 1). Research using response mode systems has attempted to describe psychotherapy process, as well as discover response modes and response mode patterns associated with positive outcome in psychotherapy. The early research by Snyder (1945) and Strupp (1955) attempted to describe the differences in therapeutic orientations according to verbal response patterns used. More recent research by Stiles (1979), Hill et a1. (1979), and Elliott et a1. (Note 4) has confirmed these early findings. For example, therapists in the client-centered tradition generally use reflections, reassurances, and information responses, 18 while avoiding advisements, interpretations, questions, and disclosures. Therapists using gestalt therapy use just the opposite pattern of responses. Therapists from the psychoanalytic tradition use a broad array of responses, including: reflections, interpretations, questions, re- assurances, disclosures, information, and advisements. In contrast to the findings that therapists of differ- ent orientations systematically use different response mode patterns, a significant study by Stiles and Sultan (1979) found that client response mode patterns are much more consistent across different clients and different psychotherapy. They also found that clients use disclos- ure and edification responses far more than any other response. In Stiles' system, a disclosure response is one in which the speaker reveals something about his own internal experience or point of view, while an edification response is a response expressing what the speaker believes to be objective information (Stiles, 1978). Given the results of the preceding study, Stiles and Sultan (1979) hypothesized that a common ingredient of positive outcome in psychotherapy may be found in the clients' verbal responses. This is especially important given the fact that therapist responses are not consistent, and a common ingredient of psychotherapy process is un- likely to be found in therapists' verbal responses. The impact of this statement can be somewhat misleading, 19 however. A common pattern of verbal responses not being found across therapists does not mean that particular response modes may not be discovered to be more helpful than others. It also does not mean that a common ingred- ient in psychotherapy process does not exist in therapists' behavior. Elliott et al. (1982a) recently found that not only is client experiencing associated with positive out- come in therapy, it is directly associated with therapists' direct reference to that experiencing process and modeling that process. The results suggest a circular process of mutual influence between the client and therapist. These data are especially interesting in light of Gendlin's (1973) theory that the therapist's behavior is only important in producing a positive relationship, and in stimulating the client's experiencing process. The therapist may do most anything with equal effectiveness as long as these basic conditions are established. The general trends for effective psychotherapy developed from the process research data offer some validation for these ideas. Even though client response patterns hold the most promise for finding a common effective ingredient in therapy using verbal response mode research techniques, use of therapist verbal responses should not be dismissed in looking for common therapist features. Stiles et a1.(1979a) attempted a first test of his hypothesis that client responses contain a common 20 effective ingredient across therapies by looking primarily at client disclosure responses. These responses were the most frequent in his previous study (Stiles and Sultan, 1979), and the most logical candidate since disclosure is the best insight response in his system. Stiles found a correlation of .58 (p<.001) between client disclosures and ratings using the Experiencing Scale (Gendlin et a1., 1968), offering evidence that disclosure responses are associated with good therapy process. A further test of the hypothesis was attempted by a student of Stiles. McDaniel et a1. (1981) attempted to show that use of client disclosure in psychotherapy is positively correlated with successful outcome. Data from the Vanderbilt Psychotherapy Project (Strupp and Hadley, 1979) were used for the ratings. Results proved negative, however. No consistent relationship was found between client disclosure and outcome, but more distressed clients did use a higher percentage of disclosure. As has been emphasized, verbal response mode research has not been restricted to descriptive studies. Attempts have been made to understand the nature of the response ‘modes themselves as separate process variables. Research focused upon client perceptions of therapist responses has compared response modes in an attempt to discover differences in their helpfulness levels (Elliott, 1979b; Elliott et al., 1981a). These studies have employed a 21 more recent research technique, Interpersonal Process Recall (Kagan, 1975), in their attempts to discover trends in client perceptions and associations between client per- 'ceptions and client helpfulness ratings (Elliott, 1979b). To date, the results have not been extremely promising. Some interesting trends in client perceptions of therapist behavior have been noted, but only interpretations have been mildly associated with client perceived helpfulness. The lack of evidence for differences in therapist verbal responses from the client's perspective suggests different avenues for research. One such research study that has taken a different approach to looking at discrepancies between the helpful- ness values of the various response modes is Elliott and Feinstein's (1981) cluster analysis of client descriptions 0f significant change events in psychotherapy. They Obtained clients' descriptions of significantly helpful and nonhelpful therapist responses which were then sorted into categories for cluster analysis. Different types of Significant change events identified in this manner could then be described in terms of verbal response modes. Their analyses revealed seven clusters of helpful eVents (see Tables 3 and 4): new perspective understand- ing . problem-solution, clarification of problem, personal contact, and client involvement. Three of these clusters were found to have significant patterns of response modes: 22 cans uLwHH w waH mm3 aw .uw mafimmm mmB m cosz use .uoumm uoz .on pHDoB uw unmaonu cm: H mm wan mm on u.:pH:oz DH towflamou H new .umnu twp H mm comma: pano3 kHHmDuom omLB usonm xcwzu we ohms pH .uaonm unwaonu uo>oc cm: H umfiu aowumo cm uno uswaoun :u nucowao one now who 3m: m we coflumowwsm ofiu umsu some ozu Oman we opoSu Houmsaonnm vcooom onu cH .oc uanE H um£3 use .06 on umns we maflaaou uoa cam coauDHOm m.m co cowcwdo me now wcwxmm OmHm mm3 m .omcommmu onu mo was ecu um .oE ou Hmowwoa mm3 umnu annoua osu cu coausHOm m wcwuowwo mm3 Anomaonv mu "Emanoue onu wcH>H0m ou mmnomoumam o>wumauouaw m>Humusou ucowao wcwuommo uoHomcsoo mm>a0>fiw uwuwm onH "mumumSHoQSm ucmuuoa -EH o3u mum whose .annonm wcwucomoud wcwpumwou meowumwwwSm Hammad; mo>fiooou ucoHHo cofiuaaom Emanoum "m Houmsau .unwsosu m wcfi>w£ m.um£u acmuoa %Hco osu uoc on.som umnu .cwo 50% So: xawnu cmo mcooEOm 3ocx cu wows m.uw upwmm H “$53 o8 ou xomn wcfi>ww kHHmoMu "uoHomcsoo he pooumpopas wzwaoom monfiuomoc unmfiao mafivcmumnmpca "N Houmsau .maom%E usonm moms >5 mono o8 mums DH .usonm ustOSu u.cvm£ H umSu wcflfiuoEom oNHHmoH o8 oxma can uHu "maafimxm .mhm3m mHoE wcHEoomn no .mawm usonm 3oc_mcw£um80w wcfimom monwuomop ucoHHo “cowumnuwm m.ucofiao no uCoHHo usonm cowquuomaH Ho m>fiuoodmumm 3o: m nuwz ucoHHo wmpw>oum HonmcDoo o>wuooamuom 3oz "a Hmumsao mwoumsau ozo>m Hamaamm kmnoumnfiu uco>m Hammfioccoz paw Hammaom .m oHQMH 23 “mans Assam» H "mmoooua wcwmaos onu ca nowmwco oHoE oEoomn ow woumasfiwum ma quHHU uH .ufl pawn m :033 aw .ucmB zaamon H um£3 .cowumosc umnu usonm unwsonu mzmzam H: ucoEo>Ho>cH ucowau "m noumSHu .pfiwe am umououcH umon he pm: m umcu tam Choocoo um3onm mu ”uoHomcsoo CH unopwmaoo whoa wawaooon no acmuoa m mm Hoaomcsoo mo mmcmm Moummuw m ou wcHEoo monwuomop ucowao uomucoo HMGOmumm no Houmsao .hma Camuuoo m we .oofio> m.m mo ocou onu aw posonm um: umsm mowumHEHuao popCDOm omCOQmmh one .uw Ho>o umw pHDoo H ”Ewanoua onu oEooHo>o panoo H umsu km3 w mmB muonHu "Hoammcsoo Eosm wcHEoo uHOQQDm HmCOfluoEm mooCoHumaxo ucofiao oocmHSmmmom "m Houmsao .GOquHOm manflmmoa m we coauoouwp oSu Cw wcHwasu name xE pmupmum .uofiumwOu mmopw >5 wcwumaop wouumum uHu ”cofimmom onu mo msoom osu ma £UH£B Emanoum mnu mo .mmeou mcflxuoz ma o£m\oL umLB mo COwuHchop prmoao m cu MCHEoo mmnwuommp ucowau anfiowm mo cowumowmwhmao "q HoumSHo .co ucoz A.ucoov .m «Hams 24 .wcHHomH mmmoca an we m>mw uH .GOHummsoow cm wonmE mm3 z umnu uHoH H was uOH m me On Conan; HHHmon u.cmoov an “HmoHano no HmuaoEwpsm mm nOHowcsoo onn mm>Hoonmm ncmHHo o£u .nouwsHonam pcoomm ofin CH .HmaoHuocDH AHHmon possum uH .omsoamon HmGOmnom onoE m nonsmB H .oEHn nom pommmna onoB o3 nmnu pocnmocoo onoE mm3 : .onmo on Boom u.cch m was .HHdum HmCOmnom umnn nsonm meu no>o HHpnmn H: uo>Hucouum nan no po>Ho>cha mm nOHomcsoo wan mo>Hoonod ucoHHo man .nmnHH on» GH "mnoumnHonsm nocHn ume oBu onm mnosw .ucoHHo 0n HHo>Humwoc wcnpcommon mm nOHmmnsoo moanomop ncmHHu conuomom noHomcnoo o>Humwmz .nm noumsHo .3oHH man xmonn on quB u.cpr H .usonm.wcchH£n paw wanmmm mmB H uth on coHnasnnouSH cw mw3 uH: "GOHumnonxm no onSmoHomHt m.ucmHHo auHB mcHnoHnoncH no wcHnasnnoncH mm noHomcsoo moocoHnmmxm ucoHHo COHnooanmHZ "N noumsHo .wchmm mm3 H nw£3 um pomsmcoo mmB : um£u nHoH H nan .xomn nmm Son 305x u.cpr H .oE wchcmnmnmvcn u.Cmm3 m onzme amen uHoH H .COHmchoo Ho monDOm m mmB uHu "mooCoHnomxo no EoHnonm m.ucoHHo wan Ho onsuoHn oumnsoowcH cm mm; no .mpn03 wcon3 men wchs wH .wchwm mH ncmHHo um£3 Ho ucHoa man pmmmHE mm: nOHmmcsoo mHooH unOHomcsoo an nOOnmnmpcsmHE wnHHooH moanommv ncoHHo GOHumoonmmmHz "H noumDHo mnmumsHo ocm>m HsHmHmccoz A.ucoov .m mHnmn 25 .moHowG< on .GOHanoOmm< HmonOHononmm amoHnoE< oSn Ho mwcHnooE um poncmmona noamm .mopofi omcommon HaHaHosao: was Humans: snuamonnncwnm no mnmnnmcm nmumsnu .Anmmnv .a .anmumanmn cam ..m.uuonnnm aonnk .mocmpst Hwnnm> oEOm poncms H .oE aHoL pHso3 m .wcszm mmB H nmnz Boax m HH uHoH H .wcHHmoH mm3 H Bo: oaHnommw on uno3 m now waHxOOH was H: “use non onm dHos now mCOHuwnooaxo m.ncoHHo mo>Huonpona naon no .coHuoonHw wconmH .oumnuopmcH mm omGOQmon m.noHomcsoo mooaoHnoaxo ucoHHo ncoEnGHommmmHH "m noumsHo .oHnmEmnmhm Hno> mEomm wcHnn Ho menu wnnn EH conHoop < .hnmmmooo: mms nsmsosu H nms3 no .nouHm mmB H umSB wm3 aOHmHoop m anSu n.npnp H .oE ponmnnon :conmnomp: pnoB one .coHusHOm oanHHop oEOw On oEoo on cm: H oxHH nHmH aHmmw Hu ”HSHmHoscs mm unomeoomnv mHsu mooconnonxm ucoHHo ”conmmom wannHon osn an waHnnoEOm naonm xCHsu no men 0n no conmmom wCHmHos ozu ocnmnao wcHnuoEOm on on uaoHHo co onammond wannnaa noHomcsoo Ho nHammn mm nnoHEoome moonoHnoaxo ucmHHo connmuconmnooHSHmHoncD “q noumsHu A.u:oov .m oHan 26 Table 4. Graphic Display of Helpful and Nonhelpful Event Clusters Helpful Event Clusters Task 5. 1. Cognitive Affective 2. Interpersonal Relation 1. New Perspective , 5. Reassurance 2. Understanding. 6. Personal Contact 3. Problem Solut1on 7. Client Involvement- 4. Clarification of Problem 27 Table 4. (Cont) Nonhelpful Event Clusters Task (\J . . Affective Cogn1t1ve —4 4. 3. Interpersonal Relation 1. ‘Misperception 4. Unhelpful Confronta- g- 'Misdirection tion Negative Counselor Reaction 5, Disappointment 28 new perspective events contained process advisements and interpretations; understanding events contained interpre- tations and reflections; and personal contact events were characterized by self-disclosures and information. Elliott and Feinstein's analyses also revealed five clusters of nonhelpful events: misperception, misdirection, negative counselor reaction, unhelpful confrontation, and disappointment. Only the misperception events were found to have a characteristic response mode pattern of primari- ly questions and reflections. The results obtained from this type of research suggest the existence of qualitative differences between helpful and nonhelpful instances of each response mode. Consequently, qualitative analyses aimed at highlighting the differences between instances of the same response mode is a promising direction for verbal response mode research. The results of such research may contain some important factors for identifying the differences between effective and ineffective psychotherapy. Currently, Elliott (personal communication, 1983) is working on just such qualitative analyses of collections of helpful and non- helpful instances of each response mode in his system. The use to which verbal response mode systems may be put are many and varied. They continue to offer an effective and useful system for describing the process of psychotherapy (e.g., Hill et a1., 1983). Their importance 29 as separate variables to be studied as part of psycho- therapy process continues to be emphasized with new approaches to research. Given continued results that identify effective uses of therapist response modes in psychotherapy, verbal response mode systems could have important influences in clinical or counseling training programs. Recent research on this topic suggests that graduate students in clinical training programs respond differently in therapy than nonclinical students, but clinical students do not change their response style over the course of their training (Shiffman, 1981). Students who participated in a ten—week workshop on response mode use, however, did change their response style to one distinctly different from other clinical trainees (Elliott et a1., 1981b). Identifying effective therapist response modes com- bined with identified patterns of effective client res- ponse mode use would have important implications for on-going assessment of psychotherapy. Therapists could relatively easily assess the progress of therapy given their clients' patterns of responses, and could identify their own weaknesses in any given therapy session based on their own verbal responses. Besides having implications for psychotherapy and psychotherapy process research, verbal response modes can continue to have an influence on communication skills of 30 adults in the general population, and on couples in need ofimproved communication. Training packages, such as Goodman's SASHAtapes (1979), can reach a substantial seg- ment of the population with minimal costs and demands. In this manner, the nUmbers of people who need these skills may be reached without putting even greater demands on already overburdened professionals. THE PRESENT STUDY: HYPOTHESES The present study is an exploratory attempt to describe trends in verbal response mode use in groups of successful and unsuccessful cases of insight—oriented psychotherapy. The exploratory nature of the study allows for a base of quantitative data to be established on the cases in use in the form of therapist and client verbal response mode ratings used for comparisons across groups, as well as attempting to identify significant verbal response modes used by both therapists and clients in insight-oriented psychotherapy. Considering these issues as separate features of the entire data analysis, there are then two primary components of this study with dis- tinct goals and hypotheses developed around each component. The first component focuses on comparing the fre- c11.1encies of therapist and client response mode use in successful and unsuccessful cases of insight-oriented Psychotherapy. These cases are also broken down into Pathological and nonpathological groups to facilitate Comparisons between groups and attempt to determine dif- ferences due to pathology. Four hypotheses are offered for the first component of the study: 31 32 1) Comparisons of response mode frequencies between therapists in the successful and unsuccessful groups will be insignificant. Stiles (1979), Hill et a1. (1979), and Elliott et a1. (Note 4) have recently emphasized the differences in response mode use between therapists of different therapeutic orientations. This tends to suggest, 1coupled with the fact that each orientation has been (effective, that the pattern of response mode use is not as ianortant an issue when considering therapist responses. fFlie therapeutic orientation is similar for the therapists 2111 this study, and an attempt will be made to equate any c1j1fferences across groups. Given this fact, the propor- tszon of response mode use should remain consistent across groups of cases. 2) Stiles and Sultan's (1979) study revealed the Llsse of more disclosure and edification responses by clients Eicrross differing therapist orientations. From this finding, tZIley hypothesized that client disclosure responses (the Dnc>st insight-oriented responses in Stiles' system) may I>Isovide a common key to positive outcome in psychotherapy. jPIliS study accepts the general hypothesis that client EXperiencing and insight are linked to positive therapeu- tlicoutcome, and suggests the second hypothesis that the frequency of insight-oriented client responses will be higher in the successful cases as opposed to the unsuccess- 2E111 cases. To avoid confusion with specific response 33 ‘mode categories, the client responses most associated with client experiencing and insight in this study--description, experiencing, exploration, and insight-~will hereby be labelled as client uncovering responses. 3) McDaniel et a1. (1981) discovered that more dis- turbed clients used more disclosure responses; consequent- ly; the pathological group will have a higher frequency (3f client uncovering responses than the nonpathological group. The second component of the study involves a step— vrise multiple regression analysis of therapist and client \werbal responses within each group. The attempt here is tn: move beyond an exploratory comparative analysis and :identify therapist response modes most associated with client uncovering responses. However, the comparative feature will remain as part of the component. Two hypotheses are offered for the second component of the study: 1) The most intuitively evocative therapist respon- ses, reflections and interpretations, will be most asso- ciated with client insight responses. A case study by Hill et a1. (1983) using a sequential analysis methodology does not support this conclusion. However, Hill et al.'s study used_a different method of unitizing client responses than that being used in this study, and the intuitive hypothesis that evocative therapist responses should elicit 34 client uncovering responses remains. 2) A greater frequency of these evocative responses ‘will be associated with client uncovering responses in successful cases as opposed to unsuccessful cases. This lnypothesis makes the assumption that a qualitative differ- :ance exists between evocative therapist responses in suc- c:essful cases and those in unsuccessful cases, thus offer- iJJg one suggestion for the difference in therapeutic outcome . In summary, the five hypotheses proposed in this s1:udy are listed as follows: Hypothesis 1: Comparisons of response mode frequen- cies between therapists in the suc- cessful and unsuccessful groups will be insignificant. Hypothesis II: The frequency of client uncovering responses (description, experiencing, exploration, and insight) will be higher in the successful cases as opposed to the unsuccessful cases. Hypothesis III: The pathological groups will have a higher frequency of client uncovering responses than the nonpathological groups. Hypothesis IV: The most intuitively evocative therapist responses, reflections and Hypothesis V: 35 interpretations, will be most asso- ciated with client uncovering res- ponses. A greater frequency of the evocative responses (reflections and interpre— tations) will be associated with client uncovering responses in suc- cessful cases as opposed to unsuc- cessful cases. METHOD Therapy Cases This study used therapy cases collected over a two year period (September, 1978 through June, 1980) for research purposes at the Michigan State University Psycho- logical Clinic, an outpatient clinic serving non-student members of the Michigan State University community and surroundings. The Clinic is a training and research agency of the Department of Psychology, and serves as a low cost clinic to adults, children, and families. The original data collection was based on the premise of obtaining rel- atively nonintrusive information on clients coming to the Clinic, and consists of pre- and post-therapy written measures as well as audiotapes from selected sessions. The purpose of the data collection was to provide a data source for research on the process and outcome of psyChotherapy. Over the time period from which the cases used in this study were taken there were approximately 115 intakes and 69 terminations of adult clients at the Clinic. Of the 69 terminations, there were 19 post-therapy therapist ratings of outcome for which no post-therapy client ratings were available. This left only 50 cases that met the 36 37 requirements of having both therapist and client ratings of therapy outcome available for use. For the purposes of this study, the 50 usable cases were classified into successful and unsuccessful outcome groups based on the post-therapy therapist and client out- come ratings. The cases were further broken down into pathology and lack of pathology groups based on the therapist and client ratings available on the SCL-90A and SCL-9OR symptom checklists. This allowed for the control of pathology as a confounding variable in the study. Disregarding cases where the therapist and client did not agree on the presence of pathology, 28 cases were finally available for possible use. These cases were split evenly between the pathology and lack of pathology groups. In the pathology group, there were five success- ful cases, five unsucceszul cases, and four cases where the therapist and client disagreed on therapy success. In the lack of pathology group, there were eight successful cases, three unsuccessful cases, and three cases where the therapist and client disagreed on therapy success. This pattern of case classification suggested the use of twenty total cases, ten across both categories of classification. Table 5 shows the final breakdown of cases for use in this study. Five successful and unsuccessful cases were chosen in both the pathology and nonpathology groups. The cases 38 in which the therapist and client disagreed on therapy success were considered unsuccessful for the purposes of this study (the therapist ratings were always unsuccessful and client ratings were always successful in these cases). Two of these cases needed to be used in the nonpathology group; consequently, two cases were also used in the path- ogy group. Cases were generally selected for successful and unsuccessful groups based on the following criteria: 1) the level of success or lack of success suggested by the outcome ratings (the most successful and most unsuc- cessful cases considered first); 2) the number of sessions per case (balance trying to be achieved between groups); and 3) the matching of therapists between groups whenever possible (allowing for greater control of therapeutic orientation). Clients The twenty clients whose therapy cases were used in this study agreed to allow the data to be used for research purposes. The clients were all community members who accepted the offer of therapy at the Clinic, and who com- pleted all of the pre- and post-therapy forms correctly. The limited number of cases available for use disallowed the control of such client variables as sex, age, and specific diagnosis. Although the possible effects of the limited control of confounding variables in this study are unknown, past research has indicated that such client .oHumHumnm H poHHmn o3n man Ha mo.v .m .wcHumn AdmnoSnonm EonH pommonomp mHucmoHHchHm wcHumn xmmnonuumomm 0N.Nm om.mm oq.qo o~.mo mqo.H¢ om.mq mm.mm oq.mm andLH uwomHu onmne onQHo umomze umoaHo mnasa onaHu mwcHumm Hononan mwmno>< mwaHnmm mononumm mwmnm>< flimmmwfl SH muons N.MN M N H q w.mq N m O m %NOHC£UNW mconmom H0 m 2 m S mCOHmmom H0 m z m z nonEDZ cmoz uCoHHU umHmmnwLH noafisz cwoz ucoHHo uwHamnmSH Q. Nm.om om.¢m .Nm.N¢ OH.mq qn.oq om.mq qm.oq oq.¢¢ 3 . andnH nmoaHo sneak onaHo umomcH unoaHu mnanH anHU mwcHumm >boHo£umH ommno>< mmcHumm HonocumH owmno>< HH msomm H msono o.mH N m N m c.HN N m N m . Honosnmmcoz mconmom H0 m S H z mGOmemm H0 m z m z nonesz cmoz ncoHHo uwHamnoLH nmnEDZ 5mm: ncoHHo anamnoSH HsmmmmooDmCD HDHmmmooDm .mwcHnmn Hononumm owmnw>m pnm .mCOHmmom Ho nonenc cmmE .xom "manonw Ho COHannome ”n oHQMH 40 variables do not seem to have a significant effect on client verbal response mode patterns (Stiles and Sultan, 1979). Therapists The thirteen therapists whose cases were used in this study all consented to take part in the Clinic's research. The group included second—year clinical psychol- ogy graduate students who were beginning their first practicum to advanced clinical psychology graduate students with more than one year of experience. The limited number of cases available for use disallowed the control of such therapist variables as sex, age, and specific therapeutic orientation. As in the client group, the possible effects of the limited control of therapist variables in this study are not completely known. Past research indicates that therapeutic orientation does have a significant impact on therapist verbal response mode patterns (Stiles, 1979). For this reason, the therapy cases used in this study were generally labeled as being insight-oriented, and results generalized to insight-oriented psychotherapy as a whole rather than to specific therapeutic orientations. Instruments A. Measures of Pathology 1. Hopkins Symptom Checklist (SCL-9OR, Derogatis et a1., l976)-—Client form (Appendix C). This measure 41 consists of ninety statements of problems. The problems comprise and load on nine symptom dimensions and a global severity index. The symptom dimensions are somatization, obsessive-compulsiveness, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism. Clients were instructed to check those statements that were current problems for them and to rate the degree of distress (0-4) associated with each problem. Level of pathology was partially determined for this study by converting the raw score of the global severity index into a T-score and accepting a T-score of less than 50.00 as evidence for lack of pathology and a T-score greater than 50.00 as evidence of pathology (see T-score norms, Appendix G). Recent research on the valid- ity of the SCL-9OR suggests that it is a valid measure of client distress from both a client and therapist perspective (Filak, 1982). 2. Hopkins Symptom Checklist--Therapist form (Appendix D). This form (SCL-90A) consists of nine symptom dimensions and one global pathology index. Clinicians were asked to rate the degree of symptomatology present (0-6) on each of the ten scales (computerized version of the SCL—90A used by the Clinic). Level of pathology was partially determined for this study by con- verting the raw score of the global pathology index into a T—score and accepting a T-score of less than 50.00 42 as evidence for lack of pathology and a T-score greater than 50.00 as evidence of pathology (see T-score norms, Appendix G). Level of pathology was finally determined by using the T-scores of both the client and therapist. Pathology was considered to be present if both T-scores were greater than 50.00, and not present if both T-scores were less than 50.00 (all T-scores used to determine presence of global pathology were taken from pre-therapy ratings only). B. Outcome Measures 1. Posttherapy Client Questionnaire (Appendix E). A 56-item client form (Strupp, Lessler, and Fox, 1969, shortened version) was given to clients at the termination of therapy. This form tapped the clients' subjective beliefs about the effectiveness of their therapy. Four questions which appeared to be the best representatives of overall therapy outcome were selected for use in this study (questions 3, 4, 11, and 15). Level of success was then partially determined from the client ratings on these four questions. For example, question 3 is, "How much have you benefited from your therapy?" The client can answer: 1) a great deal, 3) a fair amount, 5) to some extent, 7) very little, or 9) not at all. If the client answered 1 or 2, it was considered that on the basis of this question his/her therapy was successful. If the client answered 7 or 9, it was considered that on the 43 basis of this question his/her therapy was unsuccessful. If the ratings of three of the four questions clearly fell in the successful or unsuccessful range and the rating of the fourth remained neutral, then level of success was considered to be in the direction of the three common ratings. 2. Posttherapy Therapist Questionnaire (Appendix F). A 33-item therapist form included ten questions from the SCL-9OA and 23 questions relating to the therapist's subjective belief about the effectiveness of therapy (Strupp, Lessler, and Fox, 1969, shortened version). Three questions which appeared to be the best represent- atives of overall therapy outcome were selected for use in this study (questions 22, 27, and 29). Level of success was then partially determined from the therapist ratings on these three questions using the same process as that used for the client ratings. The overall level of success for a particular case was then determined using the ratings from both the client and therapist. If they agreed on the success or lack of success for the therapy, the case was grouped as such. Disagreements were considered to be unsuccessful cases for the purposes of this study. C. Process Measures 1. Therapist Response Mode Rating System (Elliott et a1., 1982b--Appendix A). This system consists of eleven categories of verbal responses commonly used by 44 therapists and other psychological helpers. The categories are described in detail using subtypes for more specific classification and using examples of possible therapist/ client verbal exChanges. The eleven categories are closed question, open question, general advisement process advise- ment, reflection, interpretation, reassurance, disagreement, self-disclosure, general information, and other. Therapist verbal responses are rated on each category using a four- point confidence rating scale. Reliabilities have been established for each category using a Cronbach alpha minimum standard of .70. Research on the validity of this rating system has shown these categories to be comparable to similar categories in other verbal response mode rating systems (Elliott et a1., Note 4). 2. Client Verbal Response Category System (Hill.eta14, l98l--Appendix B). This verbal response rating system consists of nine nominal, mutually exclusive categories for judging client verbal responses. These response modes include simple responses, requests, descrip- tion, experiencing, insight, discussion of plans, explora- tion of client-counselor relationship, silence, and other. Based on a revision by Elliott (Note 3), these categories have been modified in an effort to improve the type and amount of information available from them. The revised rating system used for this study includes the following categories: simple information, agreement, disagreement, 45 request, description, experiencing, exploration of thera- peutic relationship, insight, planning, and other. Elliott (Note 3) has identified the response modes defined as client uncovering responses in this study (description, experiencing, exploration, and insight) as being most similar to the various levels of client experiencing from Gendlin et al.'s (1968) Experiencing Scale. Client verbal responses are rated on each category using a four-point confidence rating scale. Interrater agreement has been established at the 80% level for the original categories, and content validity has been adequately established (Hill, 1978). Rating Groups Ratings of all therapist and client verbal responses were made by separate sets of judges. Each set of judges consisted of four undergraduate students who volunteered to participate in the study as trained raters. Training for each rater consisted of reading the manuals for each rating system, rating practice transcripts of various therapy sessions used in previous research, and participat- ing in group meetings in which the categories for each rating system and practice transcript ratings were discussed. During the rating of therapy cases used in this study, group meetings were held every one to two weeks to discuss the rating process, especially focusing on any difficulties experienced in this process. 46 Data Sampling Procedures One hour of therapy was sampled from each of the twenty cases used in this study from which therapist and client verbal response mode ratings were obtained. For cases involving multiple sessions for which a number of audiotaped sessions were available, three twenty-minute segments were sampled from each of the first, middle, and final third of the therapy process. Research by Karl and Abeles(l969) showed that different segments of a therapy session can be characterized by markedly different content areas. Considering these results, all data segments for these cases were sampled from the middle twenty minutes of each session used. Four cases used in this study required the use of different sampling proced- ures due to their lack of available sessions and audio- tapes. Two of these cases had only two audiotaped sessions available, and ratings were made for each case from two thirty-minute segments taken from the middle thirty minutes of each session. The remaining two cases had one audiotaped session available, and ratings were made from the entire session. General Procedure For the first component of the study, each member of the rating groups individually rated either therapist or client verbal responses from the audiotapes and transcripts of the sampled data from each case. Verbal response mode 47 frequency ratings for the successful and unsuccessful groups were obtained by averaging the ratings of the data in each group across each set of raters. Statistical comparisons were made between groups across both success- ful and unsuccessful cases as well as across nonpatholog- ical and pathological cases as dictated by the hypotheses. For the second component of the study, a step-wise multiple regression analysis was utilized. The purpose of this part of the study was to determine what therapist response modes tend to evoke client uncovering responses; consequently, only therapist and client responses were included in this analysis in which a single therapist response preceded a single client response within the therapy session. Multiple regression analysis was then performed on the averaged ratings of therapist and client verbal responses modes. RESULTS Preliminary Comments Before presenting the results, there are several methodological issues that are important to note: A) The calculation of frequencies for each data set progressed in two steps: 1) verbal response ratings obtained from the four raters in both the therapist and client response mode data sets were averaged across each response unit; and 2) with averaged ratings for each unit available, all response units for a particular data set were averaged across response mode categories producing a frequency value. This frequency value represents the percentage of responses in a given data set that are classified as the verbal response mode category in question. Collapsing the ratings in this manner also allows the frequency values to be expressed as mean ratings of each category in a given data set. It is important to note that the therapist and client verbal responses were rated on each category; consequently,frequency values for a given data set will inevitably total to a value greater than 1.00. B) All reliability coefficients expressed in the results section are interrater reliabilities using the Cronbach alpha procedure. Acceptable reliability is a matter of some debate, and is generally subject to the precedent established in the given area of study. Research 48 49 using verbal response mode category systems has generally considered reliability coefficients in the .70 to .80 range to be acceptable (e.g., Elliott et al., 1982b; Hill et a1., 1983); however, there has been some precedent for the use of coefficients .i .60 as acceptable (Sachs, 1983; Kraemer, 1981). For the purposes of this study, a reli- ability coefficient .i .70 will be considered acceptable, however, frequencies and reliabilities will be reported for all categories across all data sets irregardless of the level of reliability. C) All the tables reported in the analysis in which therapist and client response mode categories are compared between groups use a difference of proportions test of significance as discussed by Blalock (1972, pp. 228-230). The difference in frequency level for a single category across two groups in any comparison was considered significant if it reached the p_: .05 level by the two- tailed test. Given the number of comparisons in this study, approximately seven of the forty-one significant differences in category frequency levels were expected by chance. Direct comparisons of frequencies were made in this study since the primary purpose of the analysis was to assess differences in the levels of occurrence of therapist and client verbal response modes across success- ful-unsuccessful and nonpathologica1-pathologica1 groups. D) Two major issues in the design of this study 50 were the use of mixed outcome cases as unsuccessful cases and the inclusion of therapy of very brief duration requir- ing the use of different sampling procedures. Two cases were included in each of the unsuccessful groups in which the client rated the therapy as successful but the thera- pist rated it as unsuccessful (according to the established criteria). Also included in the unsuccessful groups were cases of brief duration from which less than three audio- tapes were available for sampling; consequently, sampling procedures for these cases were altered depending on the availability of one or two audiotapes. Three of these cases were included in the nonpathological-unsuccessful group (Group II) and one such case in the pathological- unsuccessful group (Group IV). Comparing cases of mixed outcome with those cases considered unsuccessful by both therapist and client we discover no pattern of differences within Group IV, and no clear pattern of differences within Group II. However, the most significant differences in this analysis are found in Group II. The comparison of mixed outcome cases with unsuccessful cases in this group also represents a comparison between the three cases using alternate sampling procedures (unsuccessful cases) and the remain- ing two using the established sampling procedures (mixed cases). 51 Table 6. Comparison of Mixed Cases with Unsuccessful Cases Within Groups II and IV. Group II Ila IIb (N=273)' (N=2D‘£)' Ther. Resp. Mode Freq) Unsucc. Reli Freq. Mixed Reli. Closed Question .109** .46 .038 .52 Open Question .405** .84 .268 .93 Process Advisement .045 .69 .021 .82 General Advisement .039 .27 .011 .41 Reflection .357 .73 .470** .78 Interpretation .104 .57 .212** .58 Reassurance .054 .73 .026 .76 Disagreement .026 .37 .011 .66 Self-Disclosure .090* .83 .035 .68 General Information .033 .44 .047 .37 Other .001 .00 .004 .00 (N = 273) (N = 251) Client Resp. Mode Freq. Unsucc. Rel1 Freq. Mixed Re11 Information .166*** .64 .061 .56 Agreement .151 .81 .251** .82 Disagreement .028 .66 .017 .70 Request .048 .92 .028 .89 Description .368 .87 .406 .78 Experiencing .237 .75 .292 .84 Exploration .113** .83 .040 .63 Insight .009 .25 .054** .64 Planning .010 .62 .016 .31 Other .025 .48 .009 .00 Group IV ,IVa IVb (N=l7'§7 (N=18T)' Ther. Resp. Mode Freq. Unsucc. Re11 Freq. Mixed Rel1 Closed Question .062 .51 .064 .40 Open Question .210 .87 .407*** .90 Process Advisement .024 .80 .028 .56 General Advisement .023 .00 .039 .10 Reflection .319 .74 .360 .74 Interpretation .184 .72 .142 .68 Reassurance 120*** .80 .021 .58 Disagreement ' .019 .45 .025 .46 Self-Disclosure .093 .74 .047 .75 General Information .056 .16 .022 .07 Other .034* .00 .000 .00 52 Table 6. (Cont) (N = 217) (N = 238) Client Resp. Mode Freq. Unsucc. Reli. Freq. Mixed Reli. Information .061 .58 .136** .57 Agreement .085 .81 .152* .86 Disagreement .036 .84 .055 .86 Request .072 .88 .099 .92 Description .427 .82 .352 .85 Experiencing .352*** .80 .183 .79 Exploration .105 .88 .065 .81 Insight .027 .16 .028 .44 Planning .016 .65 .006 .54 Other .010 .00 .026 .65 p_ < .05, *‘kp < .01, ***p < .011 Observed differences are most parsimoniously explained by the fact that sampling error and measurement error increase with a reduction in data (notice the range of reliabilities within groups), and that procedural differ- ences in sampling data will probably have some effect on frequency values in process research (Karl and Abeles, 1969). For example, the significant increase in questions found in Group 11a is expected when considering the fact that more data is sampled from the first twenty minutes of a therapy session when questions are generally in' greater abundance. It must also be noted that the thera- pists in the two mixed cases in Group II are the same individual. Considering the effect differences in thera- peutic orientation can have on therapist verbal response mode data (Stiles, 1979), it is not unexpected to find significantly higher frequencies of reflections and interpretations in Group IIb. 53 Without clear patterns of differences being established, it could be more misleading to exclude data under these methodological circumstances. For this reason, all data will be included for analysis as originally presented with the added understanding that any significant results invol— ving controversial data must be given special attention. Part 1: Comparisons of Therapist and Client Verbal Response Mode Group Frequencies Hypothesis I Comparisons of response mode frequencies between therapists in the successful and unsuccessful groups will be insignificant. The results shown in Table 7 support the first hypothesis. Comparisons of therapist response mode fre- quencies between Group I and Group II produce no signifi- cant differences. Comparisons between Group III and Group IV produce only two signficant differences, both involving categories with reliabilities below the .70 established criterion. 54 Table 7. Comparison of Successful and Unsuccessful Cases Across Nonpathological and Pathological Groups. Group I Group II (N = 358) (N = 479) Therapist Response Mode Frequency Reli. Frequency Reli. Closed Question .087 .70 .079 .46 Open Question .324 .88 .347 .88 Process Advisement .025 .54 ..035 .73 General Advisement .033 .50 .027 .29 Reflection .353 .73 .405 .75 Interpretation .195 .71 .150 .60 Reassurance .059 .75 .042 .74 Disagreement .007 .41 .018 .39 Self-Disclosure .052 .78 .066 .81 General Information .043 .18 .034 .39 Other .003 .50 .002 .00 Group III Group IV (N = 461) (N = 360) Therapist Response Mode Frequency Reli. Frequency Reli. Closed Question .091 .63 .063 .46 Open Question .344 .87 .309 .90 Process Advisement .053 .82 .026 .69 General Advisement .089*** .61 .027 .07 Reflection .278 .71 .340 .74 Interpretation .131 .64 .163 .70 Reassurance .054 .65 .070 .80 Disagreement .033 .69 .022 .45 Self-Disclosure .080 .80 .070 .74 General Information .087** .49 .032 .12 Other .006 .00 .012 .00 ME < .01, mm < .001 These results suggest that, taken as collective groups, therapists within both the nonpathological and pathological groups are responding equally across successful and unsuc- cessful cases. Any differences due to diversity of 55 orientations or differences within orientations in this study are negligible for these group comparisons. As pointed out in the preliminary comments, a compari— son of mixed outcome with consensus outcome in Group 11 did produce a number of significant differences. Even though within-group differences are expected for these data, a question arises concerning the validity of the Group I versus Group II comparison. It is possible that mixed outcome cases bias the unsuccessful data so that a Group I versus Group II comparison does not represent a distinct successful-unsuccessful comparison for nonpatho- logical cases; thus, any possible differences between the two groups could go undetected. The results shown in Table 8 are presented to help resolve this methodological weakness by comparing Group I with the separate consensus unsuccessful cases of Group 11 (identified as Group Ila). The results indicate that the first hypothesis remains supported even when mixed outcome cases are removed from Group II. The only signif- icant difference that meets an acceptable level of reli- ability is the higher frequency of open questions in the unsuccessful cases. 56 Table 8. Comparison between Successful-nonpathological cases and Consensus Unsuccessful-nonpathological Cases. Group I Group Ila (N = 358) (N = 275) Therapist Response Mode Frequency Reli. Frequency Reli. Closed Question .087 .70 .109 .46 Open Question .324 .88 .405* .84 Process Advisement .025 .54 .045 .69 General Advisement .033 .50 .039 .27 Reflection .353 .73 .357 .73 Interpretation .195*** .71 .104 .57 Reassurance .059 .75 .054 .73 Disagreement .007 .41 .026 .37 Self-Disclosure .052 .78 .090 .83 General Information .043 .18 .033 .44 Other .003 .50 .001 .00 *p < .05, ***p < .001 Hypothesis II The frequency of client uncovering responses (descrip- tion, experiencing, exploration, and insight) will be higher in the successful cases as opposed to the unsuccess- ful cases. The results shown in Table 9 do not support the second hypothesis. Comparisons of client response mode frequen- cies between Group I and Group II produce only two signifi- cant differences. The significantly more frequent explor- ation response in Group II is the only difference that meets the established reliability criterion. Comparisons between Group III and Group IV produce three significant 57 differences that meet the reliability criterion. The identified uncovering responses of description, experienc- ing, and exploration are all significantly more frequent in Group IV. The insight response in Group IV also has a higher frequency, but the difference is not significant and the insight response is not reliable in this study. Table 9: Comparison of Successful and Unsuccessful Cases across nonpathological and pathological groups. Gro. up I 92222.11 (n = 400) (N = 524) Client Response Mode Frequency Reli. Frequency Reli. Information .101 .56 .116 .64 Agreement .217 .85 .199 .82 Disagreement .044 .67 .023 .67 Request .039 .87 .038 .91 Description .361 .84 .386 .8 Experiencing .273 .81 .264 .80 Exploration .039 .81 .078* .80 Insight .040 .59 .031 .62 Planning .029* .75 .011 .46 Other .011 .39 .016 .42 Gropp III Group IV (N = 571) (N = 455) Client Repponse Mode Frequengy Reli. Freqpency Reli. Information .144* .69 .100 .59 Agreement .149 .85 .120 .85 Disagreement .032 .74 .046 .85 Request .062 .91 .086 .90 Description .283 .84 .388*** .84 Experiencing .195 .82 .264** .81 Exploration .039 .83 .084** .85 Insight .018 .44 .027 .34 Tanning ' .005 .41 .010 .61 Other. .020 .14 .018 .52 *2 < .05. **2 < .01. ***2 < .001 58 These results indicate a clear reverse of hypothesis II for pathological cases in this study. Uncovering responses are significantly more frequent in unsuccessful as opposed to successful therapy. For nonpathological cases, the only difference that exists between successful and unsuccessful therapy is also in the reverse direction. Even with mixed outcome cases removed from Groups II and IV (identified as Groups IIaanuiIVa--see Table 10) the results are not changed. Table 10. Comparison of Successful and Unsuccessful cases with Mixed Outcome Cases Removed. Group I Group Ila (N = 400) (N = 273) Client Response Mode Frequency Reli. Frequency Reli. Information .101 .56 .166* .64 Agreement .217* .85 .151 .81 Disagreement .044 .67 .028 .66 Request .039 .87 .048 .92 Description .361 .84 .368 .87 Experiencing .273 .81 .237 .75 Exploration .039 .81 .113*** .83 Insight .040* .59 .009 .25 Planning .029 .75 .010 .62 Other .011 .39 .025 .48 *E < .05, **E < ,01, ***B < .001 59 Table 10.(Cont) Group III Group IVa (N = 571) (N = 217) Client Response Mode Frequency Reli. Frequengy Reli. Information .144** .69 .061 .58 Agreement .149* .85 .085 .81 Disagreement .032 .74 .036 .84 Request .062 .91 .072 .88 Description .283 .84 .427*** .82 Experiencing .195 .82 .352*** .80 Exploration .039 .83 .105*** .88 Insight .018 .44 .027 .16 Planning .005 .41 .016 .65 Other .020 .14 .010 .00 an .05. «2 < .01. my -001 Hypothesis III The pathological groups will have a higher frequency of client uncovering responses than the nonpathological groups. The results shown in Table 11 do not support this hypothesis. Comparisons between Group I and Group III produce signficantly higher frequencies of description and experiencing responses in the nonpathological cases, suggesting a reverse of the hypothesis to be true for the successful cases in this study. Comparisons between Groups II and IV reveal no significant differences in the frequencies of uncovering responses in the two unsuccess- ful groups. These results, and the preceding results from hypothe- sis II, indicate that Group III is differentiated from the 60 other groups by a lower frequency of uncovering responses. All other groups are approximately equal in the level of these responses. Considering the small number of cases composing each group in this study, individual case Table 11. Comparison of Successful and Unsuccessful Cases Between Nonpathological and Pathological Groups. Group I Group III (N = 400) (N = 571) Client Response Mode Frequency Reli. Frequengy Reli. Information .101 .56 .144 .69 Agreement .217** .85 .149 .85 Disagreement .044 .67 .032 .74 Request .039 .87 .062 .91 Description .361** .84 .283 .84 Experiencing .273** .81 .195 .82 Exploration .039 .81 .039 .83 Insight .040* .59 .018 .44 Planning .029** .75 .005 .41 Other .011 .39 .020 .14 Group II Group IV (N = 524) (N = 455) Client Response Mode Frequency Reli. Frequency Reli. Information .016 .64 .100 .59 Agreement .199*** .82 .120 .85 Disagreement .023 .67 .046 .85 Request .038 .91 .086** .90 Description .386 .83 .388 .84 Experiencing .264 .80 ..264 .81 Exploration .078 .80 .084 .85 Insight .031 .62 .027 .34 Planning .011 .46 .010 .61 Other .016 '.42 .018 .52 *p< .05, **p< .01, ***p< .001 61 analyses to show the frequency range of each client response mode in Group III is warranted. The results of this analysis are shown in Table 12. These results show a relatively wide range of frequency levels across individual cases, and suggest a broad range of individual case variation in the use of client response modes in this group. Three of the five cases produce frequencies of at least one uncovering response that would not be signficantly different from GrOUps I and IV, Table 12. Individual Case Frequencies of Client Response Modes in Group III. Case Number Freq. Range Client Response Mode l 2 3 4 5 Min. - Max. Information .088 .213 .098 .193 .113 .088 - .213 Agreement .188 .105 .170 .101 .182 .101 - .188 Disagreement .062 .058 .023 .014 .016 .014 - .062 Request .018 .045 .097 .038 .094 .018 - .097 Description .370 .213 .364 .363 .185 .185 - .370 Experiencing .355 .217 .260 .126 .109 .109 - .355 Exploration .023 .055 .040 .066 .031 .023 - .066 Insight .031 .019 .037 .013 .010 .010 - .037 Planning .022 .004 .014 .005 .000 .000 - .022 Other .010 .000 .000 .000 .058 .000 - .058 but three of the five cases also produce frequencies of at least one uncovering response that would remain signifi- cantly lower than Groups I and IV. The results indicate that any interpretation of the differences between groups involving Group III should be made with caution. 62 Part II. Multiple Regression with Therapist and Clienthesponse Modes Data Analysis A step-wise multiple regression was performed in an attempt to explore the possible prediction of client res- ponse modes from therapist response modes. All data was averaged across the four raters in each of the client and therapist data sets, and regression analysis was performed on the data with each therapist response preceding each client response. Some response units were excluded from the analysis due to the lack of this therapist response/ client response ordered relationship in the data. The step-wise multiple regression analysis was perform— ed on each of the four groups separately to allow for comparisons. The results are reported for each client response mode category separately, and only significant (p < .05) therapist response predictors are listed. The following information is reported for each therapist response predictor: analysis of variance F-value, signif- icance level, squared multiple correlation coefficient, change in the squared multiple correlation coefficient for each predictor variable representing the amount of the variance explained by that variable, and the simple correlation coefficient providing the direction of the 63 relationship between the therapist predictor variable and the client response mode dependent variable. Group I (N = 340) Client response mode: Information Therapist Predictor Variable F Sig. R_ R ch. 3 Open Question 72.349 .001 .176 .176 .420 Closed Question 36.151 .001 .256 .080 .378 Interpretation 5.508 .020 .268 .012 -.324 General Information 7.669 .006 .284 .016 -.087 Client response mode: Agreement Therapist Predictor Variable F Sig. R: chh. 3 Reflection 53.582 .001 .137 .137 .370 Open Question 10.213 .002 .162 .025 -.336 Reassurance 5.395 .021 .175 .013 -.104 Self-Disclosure 5.966 .015 .190 .014 -.075 Interpretation 4.617 .032 .201 .011 .308 Client response mode: Disagreement Therapist Predictor Variable F Sig. R: chh. g Disagreement 6.823 .009 .020 .020 .141 Reflection 5.258 .022 .035 .015 .128 Client response mode: Reqpest Therapist Predictor Variable F Sig. R: chh p Closed Question 4.324 .038 .013 .013 -.112 Client response mode: Description Therapist Predictor Variable F_ Sig. R: chh. E Self-Disclosure 6.591 .011 .019 .019 -.138 Client response mode: Experiencing No significant relationships. 64 Client response mode: Exploration Therapist Predictor Variable F Sig. Self-Disclosure 22.238 .001 Open Question 4.247 .040 Client response mode: Insight No significant relationships. Client response mode: Planning Therapist Predictor Variable F Sig. General Advisement 16.466 .001 Reassurance 4.542 .034 Client response mode: other Therapist Predictor Variable F Sig. Self-Disclosure 33.299 .001 Other 11.131 .001 Process Advisement 4.242 .040 Group II (N = 466) Client response mode: information Therapist predictor Variable F Sig. Open Question 89.585 .001 Closed Question 21.521 .001 Process Advisement 11.630 .001 Interpretation 5.071 .025 Reflection 4.513 .034 Client response mode: Agreement Therapist Predictor Variable E Sig. Open Question 84.478 .001 Reflection 23.012 .001 Process Advisement 14.581 .001 Interpretation ' 8.792 .003 Reassurance 5.154 .024 Self-Disclosure 4.890 .028 .062 .073 I130 pd .162 .199 .219 .227 .235 R .154 .194 .219 .233 .242 .250 .062 .012 2 R ch. .162 .037 .020 .009 - .008 - .154 - 2025 - Zoos - .008 — IH .248 .085 IH .216 .110 IH .299 .185 .120 E .402 .271 .092 .306 .328 IN .392 .386 .140 .324 .090 .107 65 Client response mode: Disagreement Therapist Predictor Variable F Interpretation 14.135 Closed Question 4.386 Process Advisement 4.645 Client response mode: Request Therapist Predictor Variable F Self-Disclosure 56.750 Reassurance 4.287 Client response mode: Description Tpggapist Predictor Variable F Self-Disclosure 22.150 Client response mode: Experiencing Therapist Predictor Variable F Self-Disclosure 11.310 Open Question 4.549 General Information 5.890 Client response mode: Exploration Therapist Predictor Variable F SUlf—Disclosure 19.634 Open Question 9.537 General Information 8.385 Disagreement 5.682 Client response mode: Insight Therapist Predictor Variable F 7.055 Interpretation Client response mode: Planning No significant relationships. .001 .033 .016 Sic .001 .002 .004 .018 .008 W .030 .039 .048 .109 .117 {SJ .046 i: lw N N p~ SH .172 .066 .074 IH .330 .127 IN .213 I H .154 .118 .061 IH .201 .109 .141 .150 IN .122 66 Client response mode: Other Therapist Predictor Variable F Sic. ___£L Process Advisement 56.000 .001 General Advisement 13.232 .001 Group III (N - 427) Client response mode: Information Therapist Predictor Variable F Sig. Open Question 69.222 .001 Closed Question 10.855 .001 Other 10.531 .001 Reflection 4.371 .037 Client response mode: Agreement Therapist Predictor Variable F Sig. Open Question 50.878 .001 Reflection 16.486 .001 General Information 11.891 .001 Interpretation 7.632 .006 Client response mode: Disagreement No significant relationships Client response mode: Request Therapist Predictor Variable F Sig. Disagreement 22.648 .001 General Advisement 7.254 .007 Interpretation 5.612 .018 Self-Disclosure 5.231 .023 Client response mode: Description Therapist Predictor Variable F Sig Self Disclosure 18.223 .001 Closed Question 12.105 .001 Process Advisement 10.932 .001 General Information 4.352 .038 7d .108 .132 70 .140 .162 .182 .190 W .107 .140 .164 .179 .051 .067 .079 .090 .041 .068 .091 .100 ii C 11 .051 .016 1011 R ch .041 .027 .023 .009 .329 .241 IN .374 .215 .105 -.276 i"; —.327 .304 .269 .290 .225 .145 -.057 .146 -.203 .195 .160 -.l63 67 Client response mode: Experiencing Therapist Predictor Variable F Sig; Closed Question 5.114 .024 Self-Disclosure 4.027 .045 Other 4.090 .044 Client response mode: Egploration Therapist Predictor Variable 2 gig; Reflection 11.639 .001 Client response mode: Insight No significant relationships. Client response mode: Planning No significant relationships. Client resppnse mode: Other Therapist Predictor Variable F Sig; General Advisement 52.881 .001 General Information 10.859 .001 Interpretation 5.768 .017 Grogp IV (N = 352) Client response mode: Information Therapist Predictor Variable F Sig; Open Question 97.861 .001 Client response mode: Agreement Therapist Predictor Variable F Sig; Reflection 54.602 .001 Interpretation 7.602 .006 $15 .111 .133 .145 .111 .022 .012 ch. ['1 .109 .077 .100 I"! .163 IN .333 .278 .223 IN .467 [H .367 .294 68 Client response mode: Disagreement Therapist Predictor Variable F Sig; Disagreement 5.743 .017 Client response mode: Request No significant relationships. Client response mode. Description Therapist Predictor Variable E Sig; Reassurance 13.859 .001 Interpretation 7.315 .007 Client response mode: Experiencing Therapist Predictor Variable F Sig; Reassurance 12.124 .001 Client response mode: Exploration Therapist Predictor Variable F Sig; Self-Disclosure 6.628 .010 Client response mode: Insight Therapist Predictor Variable F Sig; Reassurance 4.789 .029 Interpretation 5.580 .019 Client response mode: Planning Therapist Predictor Variable F, Sig; Reassurance 4.718 .031 Client response mode: 'Other Therapist Predictor Variable F, Sig; Other 7.155 .008 General Advisement 3.892 .049 70 .016 .038 .058 71 .033 w .019 chh. .016 .033 .020 .011 1'1 .127 IN .195 .168 IN .183 I"! .136 IH .111 .105 IN .115 E .142 .117 69 Hypothesis IV The most intuitively evocative therapist responses, reflections and interpretations, will be most associated with client uncovering responses. The results from the multiple regression analysis do not support this hypothesis. A weak relationship between interpretation and insight was discovered in Groups II and IV, but the low reliability of the insight response mode renders its use in this study pointless. In fact, the only other occasions in which reflections or inter- pretations are significant predictor variables for client uncovering responses are when they have a negative rela- tionship (reflection with exploration in Group III and interpretation with description in Group IV). The therapist response of self-disclosure seems to have the most consistent impact in predicting client uncovering responses. A mildly positive relationship between therapist self-disclosure and client exploration is present in three of the groups, while a mildly negative relationship between therapist self-disclosure and client description and experiencing is present in three and two groups respectively. Hypothesis V A greater frequency of the evocative responses (reflections and interpretations) will be associated with client uncovering responses in successful cases as opposed 70 to unsuccessful cases. Considering the results emphasized by the preceding hypothesis, hypothesis V is not supported by this study. The multiple regression analysis does not reflect any qualitative differences in therapist responses between groups, and as stated previously, the only indication of any kind of relationship between therapist interpretations and reflections with client uncovering responses is a very mild negative relationship. The most promising possible relationship that was discovered is a consist- ently mild association between therapist self—disclosure and client exploration. It is most accurate to state that the multiple regression analysis reveals no firm predictive relationships between therapist responses and client uncovering responses. DISCUSSION The primary focus of this study centered on the quantification of therapist and client verbal responses as a way to identify potential factors in differentiating successful and unsuccessful psychotherapy, and as a way to identify relationships between therapist responses and the use of particular client responses. More generally, the potential use of therapist and client verbal response mode category systems as research and predictive instru- ments was put to question. This section addresses the central themes of this study by specifically focusing on the five questions generated by those themes. The first component of the study, comparing group frequencies of therapist and client verbal response modes, is discussed through questions I and II. The second component, focusing on the specific predictive relationships between therapist and client verbal response modes, is addressed by question 111. Questions IV and V provide an overview of the implications of these results on the use of verbal response mode category systems and the future of psychotherapy research. 71 72 I. Is the quantification of therapist response modes useful in differentiating successful and unsuccessful psychotherapy? The early research using verbal response mode category systems (Snyder, 1945; Strupp, 1955) attempted to describe differences in therapeutic orientations according to the verbal response patterns used by therapists of the same and different schools of thought. This use of verbal response modes proved to be successful and continues to be popular (Stiles, 1979; Hill et a1., 1979; Elliott et a1., Note 4; Lee and Uhlemann, 1984), but does not address the more specific issue of whether frequency patterns within a given therapeutic orientation can be identified with increased success rate. Due to a lack of systematic control over therapeutic orientation in the present study, this question was originally thought to be beyond the limits of current investigation. Any differences that would arise between successful and unsuccessful groups could more parsimon— iously be explained by orientation differences. Upon further consideration, however, an assumption was made that the graduate student therapists in this study, drawn from.the same agency and the same clinical psychology training program, would work with similar therapeutic orientations. Individual differences would indeed exist, as they exist among experienced clinicians professing 73 allegiance to the same orientation, but taken collective- ly very similar response patterns across successful and unsuccessful groups were expected. If such results were supported by the study, then the issue of differentiat- ing success rate based on therapist response mode patterns could be addressed by the sheer fact that a lack of differentiation would exist in the data. The results do indeed support a lack of differentia— tion across successful and unsuccessful psychotherapy in this data set. This finding suggests that using frequencies of therapist response modes to differentiate between effective and ineffective psychotherapy is in— appropriate, even within a given orientation. Such a finding is supported by a large amount of empirical evidence that shows psychotherapy to be effective across diverse therapeutic orientations, and reflects the notion that quantifying therapists' verbalizations may only serve to describe and categorize therapy process--not define its quality. As a point of interest, comparing the response mode patterns used by therapists in this study to those used by established professionals who represent particular schools of insight-oriented therapy (see Table 13), we can see that therapists in this study most closely follow a dynamic orientation. It should be noted that due to differences in the representation of frequency values 74 between this study and the study from which Table 13 is taken, direct comparisons between frequency values are inappropriate. Comparisons should be made based solely on patterns of response mode use. II. Is the quantification of glient response modes useful in differentiating successful and unsuccessful psychotherapy? After discovering the verbal process of therapists from different orientations to be quite distinct rather than convergent on a common mixture of techniques as many professionals had claimed (e.g., London, 1964), Stiles (1979) hypothesized that it would be the verbal behavior of the client that would allow us to distinguish effective psychotherapy from ineffective psychotherapy. This hypothesis was supported when Stiles and Sultan (1979) discovered that client response mode patterns are much more consistent across individuals and therapeutic orientation, thus providing a common feature of psycho- therapy that could potentially be used to distinguish qualitative differences. Stiles et a1. (1979) attempted a first test of hiS' hypothesis by studying the relationship between the best insight response in Stiles' verbal response mode system (client disclosure) and ratings using Gendlin et al. 3 (1968) Experiencing Scale. A correlation of .58 75 .oDUchoou oHunonmnmnnocoxmn new mopoE mmCOQmmn Hmnno> .No-as .Ne .nnumnnonmm .Amnmnv .m .3 .mmnnum sonna mopoE pmanomonn HHHmoHumnooaek one New was mmocmnwuns Ho nmnsbz n.e ~.m n.m H.H o.o «.0 ucmsmmnmmmnn m.o m.o N.o N.o N.c N.o oHnmnoomcp m.H m.MH aw.e so.oH m.H m.s onSmOHomHo w.N o.N «H.0m aN.mN 0.0 N.o namsomH>p< aN.NH aw.wH am.mH so.wN N.N m.w GOHnmmso an.me «m.m~ am.o~ «o.NN «.0 0.0 conumumnanmucn H.N N.m N.o m.m «m.H aw.o coHnmoHHHem N.o c.o 0.0 0.0 «N.o «N.o coHuwanncoo sm.nn «n.0n N.m o.o «H.0s so.es ucmsmwemnsocxoa am.m am.w N.m H.m sq.me «H.Nm coHnooHHom ucmucH Show uconcH anon ucmucH Enom oHnHHmamosonm nHwnmou pmnmncoUuucmHHo kmnmnmSH Ho monmfi dense nom wopoo neoncH new Enom Ho mowmuGUUnom .mH oHan 76 (p <.001) was discovered, thus offering evidence that a client response mode could be associated with good therapy process. The next step was to show that client response modes (more specifically the client disclosure response) could be used not only to mark good therapy process, but to describe successful psychotherapy. McDaniel et a1. (1981) used data from the Vanderbilt Psychotherapy Project (Strupp and Hadley, 1979) in an attempt to correlate client disclosure with successful therapy outcome, but could find no consistent relationship between the two. The present study moved the basic question being asked in this section to a research methodology more conducive to providing an answer. Rather than attempting to find relationships to success rate in controversial data, this study was designed to allow for the direct comparison of client response mode frequencies between successful and unsuccessful psychotherapy. The results of this study indicate that client response modes associated with client experiencing in psychotherapy do not provide a measure of therapeutic effectiveness. In fact, greater frequencies of client uncovering responses were present in unsuccessful cases in which pathology level was greater than in successful cases. Stiles' hypothesis that client response modes are central to describing effective therapy leading to greater success is not 77 supported. This finding is important not only as additional information for psychotherapy process measures, but also as part of a conceptual view of human behavior change. The assumption that was being made in this study and in Stiles' hypothesis was that a client's increased use of certain quantifiable response modes is associated with an increase in what may be called client experiencing. This experiencing process may best be described as the express- ion of an accurate and personal feeling about some event, situation, or thought expressed in cognitive and affective terms (Rice, 1974). Previous research shows this assumption to be largely accurate, and the results of this study provide no basis for disputing it. Conceptually, however, a problem arises when an inferential leap is made from describing client experiencing as being good therapy process to defining therapeutic success based upon its level of occurrence. Indeed, therapeutic success has been associated with client experiencing in a number of studies (see literature review), but does that imply that the experiencing process can be used to define human behavior change, or does it imply that it is a key facilitative ingredient much like warmth, empathy, genuineness, and unconditional positive regard? The results of this study suggest that client experienc- ing is a common element of psychotherapy process, but 78 further suggest that therapeutic success cannot be differ- entiated based upon the frequency with which clients engage in this process. Jerome Frank (1982) has argued that emotional arousal is essential to therapeutic change by supplying a motive power to undertake the effort and suf- fering of behavior change, by facilitating attitude change, and by enhancing an individual's sensitivity to environ- mental influences. His research has indicated that emotional arousal may facilitate the behavior change process, but seems to need something else to maintain the change. Certainly there are individuals whose self-effic- acy will be greatly strengthened by acknowledging and confronting their intense affect to the point that behavior change will not only be promoted, but sustained (Frank, 1982). The point to be emphasized here is whether it is the emotional arousal, or more for the purposes of this discussion the client experiencing, that produces the change, or is it what the individual does with his or her new understanding and self-awareness that becomes the difference between ' 'experiencing" emotion and changing behavior? Orlinksy and Howard (1978) provide a list of terms used to describe the therapy process, including such notables as growth facilitating, positive and negative transference, working through, corrective emotional experience, reciprocal inhibition, modeling, and positive 79 reinforcement. All of these represent techniques that stimulate the client's expression and experiencing of painful and frightening thoughts or feelings, but more importantly they represent activities in which a client engages as part of his or her experiencing process. Their goal is not simply to increase the frequency of affective arousal, but to provide a working relationship in which that arousal is transformed into behavior change. As Freida Fromm-Reichmann (1960) states, "The process of 'working through' is aimed, then, at changing awareness and rational understanding of the unknown motivations and implications of any singled-out experience into creative, that is, therapeutically effective, insight" (p. 142). Measuring the quantity of client response modes is not going to provide a qualitative measure of how a client's experiencing process is actively impacting on behavior change. The discovery of an increased frequency of client un- covering responses in the pathological-unsuccessful group as compared to the pathological-successful group deserves comment. Due to the small number of cases in each group and the relatively wide range of frequencies across individual cases discovered in Group III, the differences between Group III and Group IV are not very meaningful. An increased number of cases may likely show that the frequency of client uncovering responses in this group are 80 the same as all other groups. Three of the five cases had levels of description or eXperiencing responses equivalent to other groups. The most interesting feature of the comparison of Groups 111 and IV is the fact that pathology levels de- creased by the end of therapy in both groups (see Table 5). There was a more dramatic decrease in Group III (successful cases), but there was also a significant decrease in Group IV (unsuccessful cases) even though subjective ratings of therapy success were negative. At first appearance, the most immediate explanation for this discovery is that Group IV is a truncated suc- cessful group. Pathology levels simply have not decreased as far as in Group III due to fewer sessions per case in Group IV. Subjective ratings of success would then be affected by a post-therapy pathology level that, although significantly improved, remained above 50.00. The group averages would apparently suggest that if Group IV therapy had continued for as long as cases in Group III, then there would have been equivalent decreases in pathology and improved therapist and client ratings of success in Group IV. If the preceding explanation were accurate, the basic conclusions being presented in this section would at least be more tenuous for pathological cases. It could be 81 argued that given more sessions the increased uncovering responses discovered for Group IV would correspond to an even further decrease in pathology level than in Group III, and quite probably subjective ratings of success that would follow this positive change. However, a more careful analysis of the cases in Group IV reveals the confound in the data. As Table 14 shows, removing the mixed outcome cases from Group IV dramatically changes the average pathology levels presented for this group. Pathology level decreases to a smaller degree from the clients' perspective, and remains virtually unchanged from the therapists' perspective. Clients also left therapy after an average of only three sessions. The mixed cases con- tributed the significant reduction in pathology in Group IV, and also contributed to the higher average number of sessions. In fact, the mixed cases in this group appear to be closer to successful cases in terms of sessions and pathology ratings, suggesting the criteria for lack of success in this study to be too flexible. Given the fact that a higher frequency of client uncovering responses exists in the consensus unsuccessful cases of Group IV, the basic conclusions of this section remain. It may even be hypothesized that an increase in client uncovering responses contributed to the clients leaving therapy prematurely. If intense, negative affect was being focused on very quickly in therapy, the 82 Table 14: A description of consensus unsuccessful and mixed outcome cases of Group IV. Group IV Therapist Client Mean Number M F M F of Sessions 2 l 1 2 3.0 Concensus Unsuccessful Cases Average Pathology Ratings Clpre Thpre lpost Thpost 69.50 65.13 64.00 66.93 Therapist Client Mean Number M F M F of Sessions 2 O l 1 45.0 Mixed Outcome Cases Average Pathology Ratings Clpre Thpre CLpost Thpost 55.00 63. 0 43.50 42.758 (a) aPosttherapyrating significantly decreased from pretherapy ratings, p < .05 by the two-tailed t-statistic. continuation of the therapeutic process may have been too frightening and painful. for these more pathological clients. This hypothesis would suggest that a "critical limit" of client uncovering responses may exist for more pathological clients beyond which the therapy process may be adversely affected. Only further research in this area will allow more definitive statements to be made. 83 It should be noted that the conclusions presented in this section are also supported by the fact that the mixed outcome cases from Group IV have a higher frequency of client uncovering responses than Group III. Subjective success rate is worse than the successful cases of Group III, post-therapy pathology levels are not significantly lower than those of Group III, and average number of sessions are approximately the same. For these cases, a higher fre- quency of client uncovering responses was not predictive of greater success. Research that has been done with more pathological clients also seems to support the ideas being expressed in this section. The University of Wisconsin studies by Rogers and his colleagues (1967) on the effectiveness of increasing client experiencing through client-centered techniques with schizophrenic patients showed that these patients were able to become more open to expressing their feelings in interpersonal relationships and less likely to deny their experiences, but significant behavior change was not produced. These results seem to suggest that the patients' susceptibility to behavior change increased, but increased client experiencing was not sufficient to adequately complete the process. 84 III. Can individual therapist resppnse modes be used to predict and elicit particular client response modes? The thrust of this question asks whether significantly helpful therapist response modes can be identified based on their relationship to client response modes that represent good therapy process. Attempts have been made to address this issue by using such statistical techniques as sequential analysis, but efforts have proved unsuccessful. In one of the more recent attempts (see Table 15), Hill et a1. (1983) came to the conclusion that this type of analysis is ineffective because it only makes use of the immediate effects of therapists' responses. They intui- tively suggest that client responses are a product of all of the interpersonal exchanges in therapy, and that thera- pist responses may not have an immediate impact on client verbal behavior. They finally conclude that "although such sophisticated statistical techniques are becoming popular and initially looked promising for counseling research, they may not be appropriate for analyzing counseling interactions" (p. 16). The multiple regression analysis used in this study was another statistical attempt at determining helpful therapeutic response modes following the same style of methodology as sequential analysis. The results from this analysis seem to corroborate the conclusions offered by Hill et a1., and support their suggestion that these 85 Table 15. Sequential Analysis of the Immediate Effects of the Counselor Predominant Response on the Client's Subsequent Two Response Units* Client Response Modes Description Experiencing Insight Counselor response modes 1 2 l 2 1 2 Silence .42+ .36 .47++ .20 .04+ .02 Approval- reassurance .40 .48 .13 .13 .00 .02 Information .34 .46 .14 .16 .01 .00 Direct guidance .04- .12- .12 .24 .00 .00 Closed question .37+ .50+ .03-- .06-- .02 .02 Open question .35 .40 .10 .19 .02 .04+ Restatement .30 .47 .10 .18 .00 .00 Reflection .22 .42 .15 .18 .01 .01 Interpretation .23- .35 .11 .18 .01 .01 Confrontation .27 .39 .08 .18 .01 .04+ Note: 1 = first response unit following counselor response 2 = second response unit. Table figures refer to the pro- portion of client responses that occurred in response to each counselor category. Higher numbers indicate more frequent occurrence. Pluses indicate that this sequence occurred more than would be expected by chance (+ = p < .05; ++ = p < .01). Minuses indicate less-than-chance occurrence (- = p < .05; -- = p < .01). *From Hill, C. E., Carter, J. A., and O'Farrell, M. K. (1983). A case study of the process and outcome of time - limited counseling. Journal of Counseling Psychology, 39, 3-18. 86 types of statistical analyses offer little in studying therapeutic interactions. It should be emphasized, however, that one problem with the multiple regression analysis used in this study is that therapist and client verbal response units usually received more than one category rating; con- sequently, the independent and dependent variables (thera- pist and client response modes respectively) used in the analysis were confounded by multiple ratings. The effects of a single category have not been determined in this analysis since the relationships represented are not 'between pure ratings of single categories. It can be .argued that the analysis performed in this manner more arealistically represents the effects of response mode czategories in psychotherapy since discrete responses are t1ot as common as complex responses. This argument may be éiccurate and deserves further consideration, but does not address the question of what are the direct effects of at single therapist response mode on client response mode £188? In the past, sequential analysis has been used with rating systems in which response units were rated as only <>t1e category. This allowed for relationships to be drawn between discrete therapist and client categories, but it 143 questioned whether this accurately reflects the complex- itty of verbal behavior in psychotherapy. In essence, CHanoun 3 may be hidden in this type of analysis that are 87 more explicitly seen using multiple ratings. It is suggested that future attempts at statistically deriving the effects of therapist response modes on client response modes use a multiple regression analysis that only includes response units rated as single categories (within a multi- ple category rating procedure), and that identifies response units given multiple ratings as new variables in the analysis (e.g., open questions also rated as reflections, reflections also rated as interpretations, self-disclosures also rated as reassurances, etc.). Given the complexity of verbal behavior in psychotherapy and the use of multiple 'rating systems, path analysis should also be considered as .a next step in defining the relationships between response modes. IV, What are the implications of this study for the rise of verbal resppnse mode catagory systems? Research on the use of verbal response modes in I>sychotherapy has recently centered on their use as Iaredictive indicators of good therapy process. Attempts IIave been made to identify the most helpful therapist lresponse modes (Elliott, 1979b; Elliott et al., 1981a; IBIliott and Feinstein, 1981), as well as verifying the fiberapeutic significance of certain client response modes (Stiles et a1., 1979a; McDaniel et a1., 1981). Results to <1ate have been less than enthusiastic. Identifying 88 individual therapist response modes as being more helpful than others has been found to be inappropriate, and the results of the present study indicate that therapeutic success cannot be defined based on the frequency of client response mode use. It appears the use of verbal response mode category systems as measures of therapeutic success is futile. It may initially appear that this research and the studies preceding it have painted a pessimistic picture of the use of verbal response mode systems in psychotherapy research. In reality, research to date has more clearly defined their roles. Rather than trying to develop their use as outcome tools, it appears the original intent of response mode systems as descriptive measures of human interactions is the most appropriate. They have proved useful in describing differences in therapeutic orientations as well as describing student—professor and patient-physi- cian interactions (Stiles et a1., 1979b; Stiles, Waszak, and Barton, 1979). As process instruments, they may also be used to describe the progress and change occurring in psychotherapy in terms of how the style of therapist-client verbal interaction changes over the course of therapy or within sessions, not in terms of the quality or helpfulness of these interactions (e.g., Hill et a1, 1983). The quantitative descriptions of dyadic interactions may also provide a firm foundation from which qualitative research 89 can be used to define the significantly helpful or unhelp- ful elements of these interactions (e.g., Elliott, 1979b; Elliott and Feinstein, 1981). Verbal response mode systems have also been extended outside the realm of psychological research by their use in interpersonal communication training (e.g., Goodman's SASHAtapes, 1979). Focusing on the function and use of individual response modes in dyadic interactions has proved extremely valuable in increasing individuals' awareness of communication styles and the impact various responses have in verbal exchanges. Nppg: Hill et al.'s (1981) Client Verbal Response Category System was modified for use in this study by dividing the simple response category into three distinct categories--simple information, agreement, and disagreement. Reliabilities and .frequency levels for two of the three categories in this study suggest that this modification of Hill et al.'s system was reasonable. The reliability for simple information never reached the .70 criterion, indicating that this category needs to be better defined and given more attention during the training of raters. Unexpectedly, the insight response mode consistently had low reliabilities in this study, also suggesting the possible need for improvement of this category and an increased focus on its use during training. The raters in this study commented that it was often difficult to 90 separate the insight response mode from the experiencing response mode, and its low base rate caused them to be less vigilant in looking for its presence. A number of therapist and client response mode cate- gories did not reach the .70 level of acceptable reliabil- ity. The occurrence of many of these categories was simply of low base rate in the data, explaining most of the low reliabilities. However, the quality of training and use of raters who were not experienced therapists may have had some impact on reliabilities. Future research using these category systems should consider more stringent criteria for the selection and training of raters, espec- ially if all response mode categories are of particular interest. V. What are the implications for future research? This study supports the need for qualitative research efforts using verbal response mode systems. As stated previously, the simple quantification of response modes in psychotherapy does not appear to be a promising avenue of outcome research; however, the use of verbal response mode systems to identify and describe dyadic interactions can provide a base from which helpful and unhelpful therapist and client response qualities can be studied. Elliott (personal communication, 1983) has begun this type of research on therapist responses, but similar 91 research can be designed for clients' use of verbal behavior. These types of qualitative analyses would potentially be very useful for not only identifying important elements in successful and unsuccessful psychotherapy, but also for interpersonal communication training. Currently, training programs emphasize the use of certain response modes based on both clinical and intuitive evidence of helpfulness. Empirical verification of these uses would serve to increase the validity of these training programs. The curious discovery of higher frequencies of client uncovering responses in the pathological-unsuccess- ful group as opposed to the pathological-successful group suggests the possibility that clients in greater distress may not benefit as much from the client experiencing process as those clients who are not under pathological levels of distress. A number of factors may contribute to this possibility; including, more intensely negative affect experienced by the more pathological clients, a greater susceptibility of more distressed clients pre— maturely leaving therapy due to more intense suffering, and a greater reluctance of these clients to engage in an actively working relationship in which they are given the responsiblity for their own behavior change. These issues point toward the possibility of important thera- peutic differences existing in the successful treatment 92 of clients with differing levels of behavioral pathology, and research exploring this area is encouraged. Considering the results and methodological weaknesses of the present study, a replication is indicated that would control for such variables as therapist orientation, therapist experience level, therapist and client gender and age, client diagnosis, and number of sessions. In addition to controlling for these variables, an increase in total cases per group and consistent sampling procedures would be vitally important methodological improvements on this research design. As emphasized by the discussion of Question II, important improvements on this research design would also include more rigid criteria for lack of success that would eliminate the use of mixed outcome cases, and changes in pathology rating that would remain consistent with therapist and client ratings of therapeutic success. Interestingly, the therapist' post therapy pathology ratings in Group I (successful-nonpathological cases) increased rather than decreased, and these ratings were higher for Group I than for Group III (pathological cases at intake). Future research should consider the use of more objective measures of success, and include cases for which pathology ratings are more consistent with the criteria for success. It is hypothesized that the results from an improved research design will not be discrepant from the basic results of this study, but will 93 allow for much firmer and more confident conclusions to be drawn. More generally, the results of this study add to a growing body of evidence pointing toward an increased focus on the therapeutic use of emotional arousal and client experiencing rather than simply being concerned with their overall production. Gendlin's (1973) theory that the therapist's behavior is only important in producing a positive relationship and in stimulating the client's experiencing process tends to minimize the function and importance of the therapist in psychotherapy. This approach appears to be too simplistic, and future psychotherapy research would be wise not to deny the importance of therapeutic intervention and therapist responsibility in defining the elements of successful psychotherapy. SUMMARY The primary purpose of this study was to determine whether the frequency of verbal response mode category use by therapists and clients can be utilized to differ- entiate successful and unsuccessful insight-oriented psychotherapy. A second feature of the study focused on identifying therapist response modes that are most effect- ive in eliciting individual client response modes. Rather than correlating therapist and client verbal responses with measures of therapeutic success as has been attempted in the past, this study was designed to directly compare response mode frequencies between four groups of cases developed around successful-unsuccessful and nonpathologi- cal-pathological dimensions. A step-wise multiple regress- ion analysis was also employed to identify the relationships between therapist and client response modes within each group. The principle hypotheses were stated as follows: 1) therapist response modes would not be useful in differ- entiating levels of success; 2) client response modes defined as uncovering responses would be found in greater frequency in successful cases; and 3) the therapist response modes of interpretation and reflection would be most positively associated with client uncovering responses. The first hypothesis was supported, but the second and third hypotheses were not. Client uncovering responses 94 95 were not found in greater frequency in successful cases, and no clear relationships were found between therapist response modes and client uncovering responses. These results more clearly define the role of verbal response mode category systems as descriptive measures of therapy process, and suggest the importance of the therapist's role in psychotherapy as more than simply a facilitator of emotional arousal and client experiencing. REFERENCE NOTES 96 REFERENCE NOTES Eaton, T. T., Elliott, R., and Reimann, J. (1983). An analysis of the effects of different unitiz- ing systems on third party ratings of psycho- therapy. Unpublished manuscript, Department of Psychology, The University of Toledo. Comprehensive process analysis model. (1981). Lecture by R. Elliott from a course in inter- personal process analysis, Department of Psychology, The University of Toledo. Elliott, R. (1983). Revised outline of Hill's Client Verbal Response Category System. Unpublished manuscript, Department of Psychology, The University of Toledo. Elliott, R., Eaton, T. T., Reimann, J., Marik, N., and Lamprecht, M. (1981). Verbal response mode use across psychotherapeutic techniques: A comparison of psychotherapeutic techniques used in the "Gloria" films. Unpublished research, Department of Psychology, The University of Toledo. 97 LIST OF REFERENCES 98 REFERENCES Ashby, J. D., Fort, D. H., Guerney, B. B., and Guerney, L. F. (1957). Effects on clients of a reflective and a leading type of therapy. Psychological Monographs, 71:24. Baker, E. (1960). The differential effects of two psycho- therapeutic approaches on client perception. Journal of Counseling Psychology, 7, 46-50. Bales, R. F. (1950). Interaction_process analysis: A method for the study of small groups. Reading, Mass.: Addison-Wesley. Barrington, B. (1961). Prediction from counselor behavior of client perception and of case outcome. Journal of Counseling Psychology, 8, 37-42. Bergin, A. E., and Jasper, L. G. (1969). Correlates of empathy in psychotherapy: A.replication. Journal of Abnormal Psychology, 74, 477-481. Blalock, H. M., Jr. (1972). Social statistics. New York, N.Y.: McGraw-Hill. Butler, J. M., Rice, L. N., and Wagstaff, A. K. (1962). On the naturalistic definition of variables: An analogue of clinical analysis. In H. H. Strupp and L. L. Luborsky (Eds.), Research in Psychotherapy, Vol. 2. Washington, D.C.: American Pathological Association. Cabral, R. J., Best, J., and Paton, A. (1975). Patients' and observers' assessments of process and outcome in group therapy: A follow-up study. The American Journal of Psychiatry, 132, 1052-1054. Crowder, J. E. (1972). Relationship between therapist and client interpersonal behaviors and psychotherapy outcome. Journal of CounselingyPsychology, 19, 68-75. 99 100 Derogatis, L. R. (1977). The SCL-9O Manual 1: Scoring Administration and Procedures fBr Ehe SCL-90. Baltimore: Johns Hopkins University SEhooliof Medicine, Clinical Psychometrics Unit. Derogatis, L. R., Abeloff, M., and McBeth, C. (1976). Cancer patients and their physicians in the perception of psychological symptoms. Psychometrics, 11, 197- 201. Elliott, R. (1979a). Helper behavior rating stystem, 1979 version. Unpublished rafing manual, Department of Psychology, The University of Toledo, Toledo, Ohio. Elliott, R. (1979b). How clients perceive helper behaviors. Journal of Counseling Psychology, 26, 285-294. __ Elliott, R., Barker, C., Caskey, N., and Pistrang, N. (1981a). Measuring and predicting the effectiveness of helping responses: Correlates of client and therapist perceptions. Resources in Education. (ERIC/CAPS No. ED 192 229.) Elliott, R., Cline, J., and Shulman, R. (1982a) Effective processes in psychotherapy: A single case study using four evaluative paradigms. Unpublished manuscript, Department of Psychology, The University of Toledo, Toledo, Ohio. Elliott, R., Eaton, T. T., Reimann, J., Marik, N., and Lamprecht, M. (1982b). Thepapist response mode rating manual. Unpublished rating manual, Department of’Psychology, The University of Toledo, Toledo, Ohio. Elliott, R., and Feinstein, L. (1981). Cluster analysis of significantly helpful and nonhelpful response modes. Paper presented at meetings of the American Psychological Association, Los Angeles. Elliott, R., Stiles, W. B., Shiffman, S. Barker, C. B., Burstein, B., and Goodman, G. (1981b). Response modes in help-intended communication: Framework and recent research. Presented as a symposium at the meetings of the American Psychological Association, Los Angeles. Filak, J. (1982) Congruence of perception on client symp- toms and therapy outcome. Unpublished thesis, Department of Psychology, Michigan State University. 101 Frank, J. (1982). Therapeutic components shared by all psychotherapies. In J. H. Harvey and M. M. Parks (Eds.), Psychotherapy Research and behavior change, The Master Lecture Series, V01. 1. WaSHington, D.C.: American Psychological Association. Fromm-Reichman, F. (1960). Principles of intensive psychotherapy. Chicago, IIlinois: The University of Chicago Press. Garfield, S. L., and Bergin, A. E. (1971). Therapeutic conditions and outcome. Journal of Abnormal Psychology, 17, 108-114. Gendlin, E. T. (1973). Experiential Psychotherapy. In R. Corsini (Ed.), Current psychotherapies. Ithaca, 111.: Peacock. Gendlin, E. T., Beebe, J., Cassens, J., Klein, M. H., and Oberlander, M. (1968). Focusing ability in psycho- therapy, personality and creativity. In J. M. Schlien, H. F. Hunt, J. D. Matarazzo, and C. Savage (Eds.), Research in psychothera , Vol. 3. Washington, D. C.: Amefican Psychologica Association. Gendlin, E. T., Jenney, R., and Schlien, J. (1960). Counselor ratings of process and outcome in client- centered therapy. Journal of Clinical Psychologyy 19, 210-213. Comes-Schwartz, B. (1978). Effective ingredients in psychotherapy: Prediction of outcome from process variables. Journal of Consulting and Clinical Psychology, 4p, 1023:1035. Goodman, G. (1979). SASHAtapes: Self-led automated series on help-intended alternatives. Los Angeles: UCLA Extension, Department of Human Development. Goodman, C., and Dooley, D. (1976). A framework for help-intended communication. Psychotherapy: Theory, Research and Practice; 13, 106-117. Halkides, G. (1958). An investigation of four therapist variables. Unpublished doctoral dissertation, University of Chicago. Hill, C. E. (1978). The development of a system for classifying counselor responses. Journal of Counsel- ling Psychology, 25, 461-468. 102 Hill, C. E., Carter, J. A., and O'Farrell, M. K. (1983). A case study of the process and outcome of time- limited counseling. Journal of Counseling Psychology, 30, 3-18. Hill, C. E., Greenwald, C., Reed, K. G., Charles, D., O'Farrell, M. K., and Carter, J. A. (1981). Manual for counselor and client verbal response category systems. _Columbus, Ohio: Marathon Press. Hill, C. E., Thames, T. B., and Rardin, D. K. (1979). Comparison of Rogers, Perls, and Ellis on the Hill Counselor Verbal Response Category System. Journal of Counseling Psychology, 26, 198-203. Kagan, N. (1975). Interpersonal process recall: A method of influencing human interaction. East Lansing, Mich.: Coliege of Education, Michigan State Univer- sity. Karl, N. J., and Abeles, N. (1969). Psychotherapy process as a function of the time segment sampled. Journal of Consulting and Clinical Psychology, 333 207-212. Kirtner, W. L., and Cartwright, D. S. (1958). Success and failure in client-centered therapy as a function of initial in-therapy behavior. Journal of Consult- ing Psychology, 22, 329-333. Kirtner, W. L., Cartwright, D. S., Robertson, R. J., and Fiske, D. W. (1961). Length of therapy in relation to outcome and change in personal integration. Journal of Consulting Psycholqu, 22, 84-88. Kraemer, H. C. (1981). Coping strategies in psychiatric clinical research. Journal of Consultipg and Clinical Psychology, 49, 309-3i9. Kurtz, R. R., and Grummon, D. L. (1972). Different approaches to the measurement of therapist empathy and their relationship to therapy outcomes. Journal of Consulting and Clinical Psycholqu, 39, 106-115. Leary, T. (1957). Interpersonal diagposis of personality. New York, N.Y.: iRonald Press. Lee, D. Y., and Uhlemann, M. R. (1984). Comparison of verbal responses of Rogers, Shostrom, and Lazarus. Journal of Counseling Psychology, 311 91-94. 103 London, P. (1964). The modes and morals of psychotherapy. New York, N. Y.: Hoif, RinEhart and Winston. Lorr, M., and McNair, D. M. (1964). The interview relation- ship in therapy. Journal of Nervous and Mental Disease, 139, 328-331. Mainard, W. A., Burk, H. N., and Collins, L. G. (1965). Confrontation versus diversion in group therapy with chronic schizophrenics as measured by a "positive incident" criterion. Journal of Clinical Psychology, 21, 222-225. McDaniel, S. H., Stiles, W. B., and McGaughey, K. J. (1981). Correlations of male college students' verbal response mode use in psychotherapy with measures of psychologi- cal disturbance and psychotherapy outcome. Journal of Consulting and Clinical Psychology,_49, 571-582. Mintz, J., Luborsky, L., and Auerbach, A. (1971). Dimen- sions of psychotherapy: A factor-analytic study of ratings of psychotherapy sessions. Journal of Consult- ing and Clinical Psychology,36, 106-120. Mullen, J., and Abeles, N. (1971). Relationship of liking, empathy, and therapist's experience to outcome of therapy. Journal of Counseling Psychology,18, 39-43. Nagy, T. F. (1973). Therapist level of functioning and change in clients' quantifiable anxiety level and verbal behavior. Dissertation Abstracts International, 22; 878B-879B. Orlinsky, D. E. and Howard, K. I. (1978). The relation of process to outcome in psychotherapy. In S. L. Garfield and A. E. Bergin (Eds.), Handbook of psychotherapy and behavior change: An empirical analysis (2nd. Edi). New York, N.Y}: Wiley. Rice, L. N. (1965). Therapist's style of participation and case outcome. Journal of Consulting Psychology, 29, 155-160. Rice, L. N. (1974). The evocative function of the thera- pist. In D. A. Wexler and L. N. Rice (Eds.), Innovations in client-centered therapy. New Yorkg N.Y.: Wiley. Rogers, C. R. (1957). The necessary and sufficient condi- tions of therapeutic personality change. Journal of Consulting Psychology, 21, 95-103. 104 Rogers, C. R. (1980). Empathic: An unappreciated way of being. In C. R. Rogers, A way of being. Boston, Mass.: Houghton Mifflin. Rogers, C. R., Gendlin, E. T., Kiesler, D. J., and Truax, C. (1967). The therapeutic relationship and its impact: A study ofypsychotherapy with sdhizophrenics. *Madison. Wisconsin: The University of Wisconsin Press. Russell, R. L., and Stiles, W. B. (1979). Categories for classifying language in psychotherapy. Psychological Bulletin, 86, 404-419. Sachs, J. S. (1983). Negative factors in brief psychother- apy: An empirical assessment. Journal of Consultipg and Clinical Psycholqu, £1, 557-564. Saltzman, C., Luetgert, M. J., Roth, C. H., Creaser, J., and Howard, L. (1976). Formation of a therapeutic rela- tionship: Experiences during the initial phases of psychotherapy as predictors of treatment duration and outcome. Journal of Consulting and Clinical Psychology, 44, 546-555. Schauble, P. C., and Pierce, R. M. (1974). Client in-therapy behavior: A therapist guide to progress. Psychother- apy: Theopyquesearch and Practice, 11, 229-234. Scher, M. (1975). Verbal activity, sex, counselor experience, and success in counseling. Journal of Counseling Psychology, 22, 97-101. Shiffman, S. M. (1981). The effects of graduate training in clinical psychology on performance in a psycho- therapy analog. Dissertation Abstracts International, figy 2084B-2085B. Sloane, R. B., Staples, F. R., Cristol, A. H., Yorkston, N. J., and Whipple, K. (1975). Ps chothera versusBeha- vior Therapy. Cambridge, Mass.: Harvard University Press. Smith, M. L., Glass, G. V., and Miller, T. I. (1980). The benefits ofypsychotherapy. Baltimore, Maryland: The Johns HopkinsiUniversity Press. Snyder, W. U. (1945). An investigation of the nature of nondirective psychotherapy. Journal of General Psychology. 33, 193-223. 105 Stiles, W. B. (1978). Manua1_for a taxonomy of verbal response modes. Chapel Hill, N.9C.: Institute for Research in Social Science, University of North Carolina at Chapel Hill. Stiles, W. B. (1979). Verbal response modes and psycho- therapeutic technique. Psychiatry, 321 49-62. Stiles, W. B., McDaniel, S. H., and McGaughey, K. (1979a). Verbal response mode correlates of experiencing. Journal of Consulting and Clinical Psychology, 47, 795-797. Stiles, W. B., Putnam, S. M., Wolf, M. H., and James, S. A. (1979b). Interaction exchange structure and patient satisfaction with medical interviews. Social Science and Medicine, 13A, 335-341. Stiles, W. B., and Sultan, F. E. (1979). Verbal response mode use by clients in psychotherapy. Journal of Consulting and Clinical Psychology, 47, - Stiles, W. B., Waszak, C. S., and Barton, L. R. (1979). Professorial presumptuousness in verbal interactions with university students. Journal of Experimental Social Psycpplggyy 15y 158-i59. V Strupp, H. H. (1955). An objective comparison of Rogerian and psychoanalytic techniques. Journal of Consulting_ Eaishelosz. 19. 1-7. Strupp, H. H. (1980a). Success and failure in time-limited psychotherapy: A systematic comparison of two cases: Comparison 1. Archives of General Psychiatry, 37y 595-603. Strupp, H. H. (1980b). Success and failure in time-limited psychotherapy: A systematic comparison of two cases: Comparison of two cases: Comparison 2. Archives of General Psychiatry, 37, 708-716. Strupp, H. H. (1980c). Success and failure in time-limited psychotherapy: Further evidence: Comparison 4. Archives of General Psychiatry,37, 947-954. Strupp, H. H., and Hadley, S. W. (1979). Specific versus nonspecific factors in psychotherapy: A controlled study of outcome. Archives of General Psychiatry, as, 1125-1136. 106 Strupp, H. H., Lessler, R., and Fox, K. (1969). Patients view their psychotherapy. Baltimore, Maryland: The Johns Hopins University Press. Tomlinson, T. M., and Hart, J. T., Jr. (1962). A valida- tion study of the process scale. Journal of Consult- ing Psycholpgy, 26, 74-78. Tomlinson, T. M., and Stoler, N. (1967). The relationship between affective evaluation and ratings of therapy process and outcome with schizophrenics. Psychother- apy: Theory, Research and Practice, 4, 14-18. Truax, C. B. (19703). Therapist's evaluative statements and patient outcome in psychotherapy. Journal of Clinical Psychology, 26, 536-538. Truax, C. B. (1970b). Length of therapist response, accurate empathy and patient improvement. Journal of ClinicalPsychology, 29, 539-541. Truax, C. B., Altman, H., Wright, L., and Mitchell, K. M. (1973). Effects of therapeutic conditions in child therapy. Journal of Community Psychology, 1L 313-318. Truax, C. B., and Carkhuff, R. R. (1967). Toward effective counseling and psyghotherapy. Chicago, Illinois: Aidine. Truax, C. B., and Mitchell, K. M. (1971). Research on certain therapist interpersonal skills in relation to processand outcome. In A. E. Bergin and S. L. Garfield (Eds.), Handbook of psychotherapy and behavior change: An empirical analySis. New York, N.Y.: Wiley. Truax, C. B., and Wittmer, J. (1971). Patient non-personal reference during psychotherapy and therapeutic outcome. Journal of Clinical Psychology, 214-300-302. Truax, C. B., Wittmer, J., and Wargo, D. G. (1971). Effects ofthe therapeutic conditions of accurate empathy, nonpossessive warmth, and genuineness on hospitalized mental patients during group therapy. Journal of Clinical Psychology,27,7137-l42. Van der Veen, F. (1967). Basic elements in the process of psychotherapy. Journal of Consulting Psychology, 31, 295-301 APPENDICES 107 APPENDIX A THERAPIST RESPONSE MODE RATING SYSTEM 108 APPENDIX A THERAPIST RESPONSE MODE RATING SYSTEM Introduction for Users of this Manual The purpose of this rating manual is to begin to teach you to do accurate and reliable ratings of the types of communication behavior used by therapists and other psychological helpers. This rating manual was developed and adapted from the framework proposed by Gerald Goodman and incorporated in his SASHAtape training series. This version incorporates many changes which have arisen out of studying help-intended communication, training and working with raters, and discussing the response modes with colleagues and friends. The form presented here includes 11 categories and is designed for use in the slow, careful analysis ("microanalysis") of audiotape recordings of help-intended conversations. The categories are as follows: Closed Question Open Question General Advisement ("Out of Session”) Process Advisement ("In Session") Reflection Interpretation Reassurance Disagreement Self-Disclosure General information Other Rating help-intended communication is challenging work, requiring much patience, care and good listening skills. Because of the care and attention to detail required, it is not a good idea to do rating when you are tired or for more than several hours at a time. However, many previous raters have reported finding their efforts rewarded by personal gains in ability to listen and respond helpfully to others. This rating system differs from most comparable response mode rating systems in a number of ways. Three of these differences are important enough to spell out right at the outset: 109 110 (l) MultipleCflassification. This response mode system takes into consideration the observation that when people communicate, their behavior reflects multiple intentions. For example, a therapist might say, "Have you tried talking to your boss about this problem?" In responding this way, a therapist is doing two different things at the same time--gathering information by asking a Closed Question, and trying to guide the client's behavior through making a suggestion, a kind of General Advisement. This means that any given response may fit into many different categories; if you decide that the response is a Process Advisement, you still have to consider whether or not it's a Self-disclosure, a Reflection, a Disagreement, and so on. Throughout the rating manual there are notes to help you with some of the more common types of multiple classification. (2) Confidence Rating Scale. This system also builds on the 6bservation that communicational acts are not either/or events. Instead, helping responses should be graded on a continuum for each particular category (technically, they should be called "dimensions" instead of "categories"). This continuum has to do with how clearly a response counts as a case of a particular response mode. For example, Reassurances aren't simply present or absent--they are present (or absent) to some degree, or with varying degrees of clarity or obvious- ness. "That's terrific" is a clearer instance of Reassurance than "That's hard” is. For this reason, this rating system uses a 4-point confidence rating scale for rating therapist response modes. The purpose of this scale is to let you record how much evidence you feel you have for rating a response mode as present or absent. Here is the confidence rating scale you will be using. '30" "1" No evidence for mode, clearly absent Maybe present, but not enough evidence to really say it's there "2" = Probably present, although there is some doubt "3" = Clearly or definitely present, strong evidence for mode (3) Subt “es. The most important feature of this response mode system is the description of subtypes for each mode. This system takes into consideration the fact that peOple can carry out a helping intention in many different ways. For example, the manual describes 111 a variety of ways in which an interpretation can be carried out, including making causal connections, labelling the client, drawing parallels, and referring to experiences which the client may not have described outloud yet. The large number of subtypes is usually a source of anxiety to beginning raters. However, please accept the Reassurance that you will not (in most instances) be asked to become accurate at identifying all the 50-odd subtypes. In fact, the purpose of the subtypes is simply to provide more detailed behavioral descriptions of the major modes. Units. This manual does not specify the type of therapist speech unit to be used. In fact, it can and has been used with all three of the major kinds of units used in response mode research: speaking turns, verbal sentences, and clauses. (We did a study comparing units and found they made little difference to reliability or validity, although they did drastically change the over- all levels of response modes.) In actual practice, the units will either be defined for you in advance or you will be taught to define them yourself. What this Manual Doesn't Do. Just reading this manual will not make you an accurate, reliable rater. For that, you need to go through a training period of practice, feedback and discussion with other expert or novice raters. In addition, this manual does not describe other practical or technical aspects of the rating process, such as standards and methods for measuring reliability (we usually use average raters together and use Cronbach's alpha to measure reliability, with a minimum standard of .70); recommended number of raters (we find 3-4 best); suggestions for adapting or expanding categories; limitations, and so on. However, you are welcome to contact the first author if you have questions or suggestions for improvements. Acknowledgements. The Therapist Response Mode RatingiSystem (Elliott, 1982 Version) builds on a tradi- tion of research on psychotherapy and counseling which goes back more than 30 years and includes the work of Snyder (1946), Strupp (1958), Lennard and Bernstein (1961) and many others. The first version of the present system was developed in 1976 by the first author in collaboration with Gerald Goodman and David Rapkin. The initial version was followed in 1979 by a revision in- corporating the influence of Clara Hill, Manny Schegloff, Robin Drapkin, and Steve King, as well as the suggestions 112 and feedback of several sets of raters: Larry Feinstein, Susan Patton, Laura Read (UCLA); and Dennis Doyle, Lisa Goldman, Maisie Lee, Tammy Linhart, Mary Lou Longnecker, Bruce Lung, and Karen Young (University of Toledo). The present version is not a major revision of the 1979 version, but instead reflects a general tightening up of definitions and examples. In addition, we have also followed the lead of Shapiro, Irving, and Barkham's (1980) revision of the 1979 version in establishing General Information as a separate category; we have also adopted several of their examples, here used with their permission. I. QUESTION Questions are characterized by the intention of gathering information for the therapist. Although there are two major categories of questions, there are several general clues which can be used for identifying either type: (a) question toneh usually rising at the end of the sentence (e.g., "you're going home?"); (b) question word order, involving switching the position of subject and’main verb or auxiliary (e.g., "are you angry right now?"; "Do ypp remember when it first happened?"); (c) questiSn words, including "who," "what," "when," ”where," "why,“ "how," and sometimes ”do” (when it's the first word of the sentence) (e.g., "How d9 you feel about that?,” "Wh do you want to know about me?," "29 you like that?”5. A. CLOSED QUESTION. Closed questions ask for specific information, thus limiting or structuring the nature Bf’the client's response. There are three subtypes: l. YES/NO QUESTIONS request a "yes" or "no answer from the client, e.g., "Did I hear you right?" "Are you the oldest then?" 2. SPECIFIC INFORMATION QUESTIONS ask for specific factualvinformation, such as a name, place, time, age, or number. They generally require only a one- or two-word answer: "How old are you?" "How many kids were there in your family?" "What do you do for a living?" 3. CLOSED-OPEN QUESTIONS. (See under Open Question. Note: Code "1" for Closed Question, "3" for Open Question.) 113 114 B. OPEN QUESTION. Open Questions ask for information or clarification without restricting the scope of the response. Open Questions have been found to be similar in intention to Process Advisements in that they usually attempt to guide the client in the session; however, they differ in that the guidance-in open questions is restricted to getting the client to provide information or clarifica- tion. 1. WORD QUESTIONS account for most Open Questions: "Why do you think that made you so sad?" "What would you like to talk about today?” "What do you mean by self-destructive?" 2. TELL-ME QUESTIONS are direct requests for information: ”Tell me more about your family." "Please try to describe for me what it's like when you lose control." Note: These should be not coded as process advisement. 3. "I" QUESTIONS occur when the therapist refers to him/herself in order to get the client to give information: "I didn't quite follow the last thing you said." "I wonder why you did that?" Note: These should not be coded as self-disclosures. 4. CLOSED-OPEN QUESTIONS are ones which are structurally closed (they have a yes/no form), but which act semantically as open questions because they allow for a broader range of responses. They often include words like "any" or "anything": "Is anything else going on in your life right now?" "Can you remember any other experiences which were like the one you just told me about" "Do you smoke pot or something?" Note: Assign such responses a "1" for closed question and a "3" for open question. 5. FILL-IN QUESTIONS. In these, the therapist stops in the middle of a sentence (without trailing off) and leaves a definite blank intended to prompt the client to fill in the missing information: C: "Up until this week it's been OK." T: "And now you're feeling: - - -" C: "- - - very uptight about seeing him again." Exception: If any of the above subtypes involve a request for specific information, then they should be coded as closed questions (e.g., "Tell me where you were born.") ‘II. ADVISEMENT The advisement response mode includes responses in which the helper or therapist tries to uide the client pp do some action, either durin the session (Process Advise- fiEnt) or outside pf the sess1on (General Advisement). In either case, advisements must meet Egp_criteria: (a) They must describe or imply some future action by the client, and (b) They must contain some kind of "push" from the helper for the client to do some action. A. GENERAL ADVISEMENT Helpers use these responses to try to get clients to do some action outside of this or other helping sessions. Usually the act1on has f5 do wiEH the client's general problem or concern; but it can also include "homework" assignments. General Advisements may be divided into four major types, which vary in terms of their forcefulness or directness: l. COMMAND GENERAL ADVISEMENTS are imperatives: "Try to stay clear of your family while you're feeling this way." "Go out and start looking for work tomorrow." "DO it o I! 2. OBLIGATION GENERAL ADVISEMENTS state the client's dbligation to do something (i.e., "shoulds”) or give the helper's verbal preferences or judgements about something that the client might do or might want to do. They can be identified by looking for obli ation words (e.g., "should," "must," "ought to"), for value words (e.g., "good," "bad," "lousy," "irresponsiEIe," "marvelous"), or for reference words (e.g., "I want," "WiSh." "need," l'IJ'JZe," "Hope," "expect"): "You should tell her you're sorry." "You ought to think of your mother before you do that." "I wish you would see a doctor about that." 116 117 ”It would probably be a bad idea to continue the relationship." "You'll have to teach yourself how to relax before dealing directly with your insomnia." 3. SUGGESTION GENERAL ADVISEMENTS suggest the possi- bilit of the client doing something. They can 5e 1dentified by looking for phrases such as "you can." "you might," ”you may,” " would you like to," "would you be able to," "have you tried," "why don't you." They also include the use of the word "please." "You might be able to convince your parents to take their vacation some place else.' "Please consider the possibility of telling him how you really feel about the relationship." ”Why don't you stay with someone else for awhile" 4. INFORMATIONAL GENERAL ADVISEMENTS are relatively indirect Advisements which try to get the client to do something by giving the client information about (a) the consequences or an action for the client (consequence format), (b) the helper (self- disclosure format), (6) other people who have done the action (information format). These can get tricky, because sometimes either the future client action or the "push" for the client to do the action may not be stated explicitly. "You know, if you keep on drinking while driving, you're going to get into real trouble." (consequence format) "If I were you, I'd tell him to stop hassling me. (N: self disclosure format; assign a "l” for self-disclosure as well) "I was once in a similar position, and I just told myself to stop being so hard on me." (assign a "3" for self-disclosure as well) "You know, your mother probably doesn't need any more hassles right now." (assign "3" for information) "One client of mine who was artistic like you kept a detailed dream journal, which proved to be very helpful to her,” (consequent and information format; assign "3" for information) IRatin note: homework assignments. Giving the client some Spemific task to performfbefore a specific session (usually tflne next) is a specialized. type of General Advisement: "Why don't you write an autobiographical statement for me?" 118 "Keep track of your automatic negative thoughts and bring me a list of them next time." ”I want you to practice relaxing with this tape once a day for the next two weeks." B. PROCESS ADVISEMENT Process advisements are intended to get the client toch) something during the helping session. Attempts to get the client to prov1 e information or clarification should not be rated as process advisements (rate them as questionETT Process Advisements may be divided into the same four major structural subtypes described for General Advisements. In addition, it is useful to describe some content cues useful for identifying specialized types of Process Advise- ment (see rating notes, below). 1. COMMAND PROCESS ADVISEMENTS use imperative struc- ture to advise the client to do something in the helping session. "Stop bad-mouthing yourself!" ”Let's get started." 2. OBLIGATION PROCESS ADVISEMENTS state the client's Obligations, appiy value-labels to the client's behavior, or give the helper's preferences for the client's behavior within the helping session. "I need for you to talk about what happened last time." "It's not very helpful for you to wait until the end of the session before bringing up things like that." (N: Rate also as "3" Disagreement) "You've got to take responsibility for what we talk about in these sessions." 3. SUGGESTION PROCESS ADVISEMENTS offer the client possibIE‘courses of action within the helping session. "Why don't you talk about what the anger's like for you?" "One possible way to go would be for us to try a little experiment right now." 4. INFORMATIONAL PROCESS ADVISEMENTS guide the client's action in the helping session indirectly by giving information about previously discussed topics, the point of time within the helping session, 119 consequences of behavior, or the behavior of the therapist or other people. "We're almost out of time." "A while ago you mentioned that you thought you deserved all this." "Last week we were talking about how your family abused you." "Most clients find it useful to begin the session by describing what happened during the previous week.” Special content cues for identifying Process Advisements. It is useful to "flag" some specialized, hard-to- identify examples of Process Advisement using the content of the action they guide the client to do: (a) Opening/Closing Session. Some Process Advisements (b) are ritualized utterances used by helpers to begin and end sessions: "OK, why don't we begin?" "Where do you want to start today?" "I guess that's all for today; it's been nice talking to you." "We have about 5 minutes left." (Re-)introduction of topics. These Process Advise- ments rely partly on the—immediate context in the helping session: The helper brings up a new topic or reintroduces an old topic, in order to get the client to talk about it. "I noticed that you wrote something about suicide on the form you filled out." "Last time you said you wanted to talk about your plans for next year." "And then there's your family---" Note: Do not rate Open or Closed Questions (includ- ing Tell-me questions) as Process Advisements, even when they introduce new topics. (e) Exercise instructions. Many process advisements have to do with setting or running in-session "exercises": "Let's try an experiment in here; close your eyes--—" "Now imagine that you're standing on a hill." "Now, change chairs." 120 (d) Role performance. Some process Advisements describe What action is expected of the client in general within helping sessions: "The first rule is to say whatever comes into your mind and hold nothing back." "I expect you to take responsibility for telling me about any suicidal feelings you may be having." III. REFLECTION With reflection, the helper's intention is to represent or feedback the client's message in some way. In’deciding whether a response is a e ection, one must make use of contextual information and one's own empathic abilities. Two basic criteria must be set in order for a response to qualify as a Reflection: (a) Meaning match. First, the helper's response must match the meaning of what the client has said. To determine whether a meaning match occurs, you should make sure you understand the meaning of the helper's response, then scan back through the client's previous talk to see if s/he has said anything which matches the meaning of the helper's response. In general, you only need to scan the 2 previous client responses for meaning matches. The meanings of client and helper responses may match at the level of words, content, implications, nonverbal messages or broad topics of conversation (see below for a discussion of these). (b) Deliberate re-presentation. Second, if there is something in the‘heIpefTs response that matches the client's response, it must be more than just mentioned as part of some other activity (such as an open question or an inter- pretation)-~it must deliberately re-present client's message. For the purpose of identifying Reflections, it is useful to be aware of a number of Reflection subtypes. The first 5 have to do with the level or type of meaning which is matched between client and helper responses: 1. QUOTE REFLECTIONS match the client's words-- they repeat the exact words of the client. These are sometimes referred to as "echoic Reflections." C: "---and it really upsets me." T: "It really upsets you." 2. PARAPHRASE REFLECTIONS match the content of what the cIient has been saying: they re-present the 121 122 meaning of what the client has just said, but translated into the helper's own words: C: ”I know I wrote it down on that form, but I don't know if this is a good thing to talk about." T: "You're not sure any more that you should talk about what you decided to earlier." T: "You mean you're having doubts about talk- ing about it now?" (N: Classify as a closed question also) IMPLICATION REFLECTIONS match what the client has been saying "between the lines” or indirect- ly implying or hinting at. These often require careful analysis to distinguish them from In- terpretations. Implication reflections take the next logical step from what the client has been saying. Examples: C: "On the other hand, if you want to get into dental school, you have to take more chemistry." T: "Ah, so you're also considering going into dentistry?" C: "And it's always been that I had to be the strong one in the family." T: "I guess that means that there was no one there for you when you had problems.” Rating note: Because they involve guesswork, many of these should be assigned weights of "2." 4. NONVERBAL REFLECTIONS match what the client has been communicating nonverbally; They re-present the nonverbal aspects of the CIient's talk including the manner or way in which s/he has been talking: T: "You're wearing a terrible frown today." C: (very fast) "I got a great new job! I can hardly wait!" T: "I can hear the excitement in your voice." 123 C: ”(sniff) I - uh (sigh) just --- (cries)." T: "You sound very sad and hurt.” SUMMARY REFLECTIONS re-present broad topics of conversation from earlier in the helping sess- ion. Unlike other kinds of reflection, they refer to more than the previous two client responses. They review and sum up a particu- lar idea that the client has been communicat- ing, a segment of the session, or even the whole helping session: "You've covered a lot of ground today: first you talked about how e11 things were going at your new job, then you brought me up-to-date on how things were going with your wife, then that led into a discussion about ---” "For the past few minutes, you've been telling me about all the things you're gonna miss now that you're graduating." Be careful to distinguish between Summary Reflect- ions and Reintorduction Process Advisements--both refer back to things the client has said earlier in the session, but their intentions differ. The last two subtypes of Reflection are specialized forms: FIRST-PERSON REFLECTIONS occur when the helper takes the role of Ehe client and speaks in the first person as if s/he were the client: C: ”And I wonder just what I'm going to have to do to put all the different things I'm doing together." T: "How can I organize it all for myself." Rating note: Do not code these as Self-disclosure. Open Question or Process Advisement. 7. COLLABORATIVE ("FILL-IN") REFLECTIONS occur when the clientIIeaves something unsaid, either by stopping in the middle of a sentence or by leaving something out of the sentence and the therapist picks up where the client left off and finishes or re-finishes the sentence for the client: 124 "And I think that was it." "That in your hands was the power to really hurt someone else." "And I still don't know ---" "How's it going to turn out." IV. INTERPRETATION Interpretations are characterized by their intention to explain or give to the client new information about the client. They are one of the most difficult responses to identify accurately. In order for a response to be classi- fied as an Interpretation, three criteria must be satisfied. (a) "News." A meaning match test in which client's and therapist's talk are compared (see Reflection, above) shows that the therapist's response con- tains at least some information which has not been stated or implied by the client earlier in the helping session or relationship. (b) "About client." The new information is about the client or some personal aspect of the client (behavior, feelings, thoughts, attitudes, per- ceptions); that is, interpretations do not focus on third parties of the world. (N: Such responses are classified as Information.) (c) Creation of understanding. In addition to these two criteffa, Interpretations also aim at creat- ing understanding or meaning for the client (as opposed to guiding the client's behavior). There are five basic types of understanding which Interpretations deal with: l. Causal Interpretations are causal inferences; they invdlve cause-effect statements; they state reasons they identify sequences of behavior. C: "I don't know why but I've just been staying away from her." T: "Probably you're afraid to start a relation- ship with her because you've been hurt before." (cause-effect) C: "My parents and I just don't seem to be getting along." T: "Maybe the reason you're having so much trouble with them is that you're still living with them." (reason) 125 f 126 C: "I don't understand why all these people are being so rotten to me." T: "It seems like whenever you get disappoint- ed, you turn it around and blame someone else instead, then when the other person reacts negatively you feel justified." (sequence) Rating note. It's important to make sure that the therapist is not Reflecting a Self-interpretation by the client. 2. Classifying Interpretations explain by putting the Client or the CIient'sIbehavior, feelings or thoughts into a new category. In other words, the therapist names, labels, diagnoses, or in some way classifies something about the client. These labels or classifications often involve technical or psychological terms (e.g., "inferiority complex", "Underdog", "manicédepressive", or "automatic negative thought"). C: "What do you think of me?" T: "You're a nice person." T: "Actually you seem quite at peace with your- self.” C: "Afterwards I started to wonder if I did the right thing: T: "Yes, I think that behavior of yours is self- destructive;" C: "What's happening to me?" T: "It looks to me like you're in the middle of a depressive episode." Rating Note. When the label is an evaluative one (good, bad, etc.), then Classifying Interpretations should also be rated as Reassurances or Disagreements. 3. Parallel Interpretations describe similarities bétween two or more real events or situations involving the client. They are used by ther- apists to point out themes or patterns to the client. 127 ”Isn't it interesting how this issue of respon- sibility keeps coming up? Earlier you were talking about not wanting to take responsibility for the problems in your relationship with your husband, and now we're talking about how you really don't want to be responsible for what we talk about here." (Rate also as Summary Reflection) "I keep hearing a feeling of sadness and dis- appointment coming up over and over again." Predictions occur when the therapist gives the client a picture of what might happen to client in the future, but without attempting to guide the client behaviorally. "How much more therapy do you think I need?” I think you're going to need another year." C T C: "Do you think I'll be able to manage?" T- "There are going to be problems and setbacks, but I think you ll be able to make it through all this." (N: code as Reassurance also) T: "It seems to me that if you knew more about these headaches then you might be able to control them." (N: Classify as General Advisement also.) Inside Interpretations are used by therapist to describe feelings or thoughts which may be going on inside the client but which the client has not yet described out loud: C: "I don't know if this is a good thing to talk about." T: "You're probably worrying that you're wast- ing my time." C: "I just hurt my child." T: "And now I guess you're feeling like a horrible monster." C: "I feel sad." T: "I think you're still really very angry about what happened last week." 128 C: ”I'm so depressed--my parents just don't understand me." T: "Are you thinking of killing yourself?" (N: Code as Closed Question also) Rating notes: (a) Inside Interpretations differ from Implication (b) (e) Reflections only by degree: Inside Interpreta- tions involve a larger inferential leap; they do not follow logically from what the client has said before--an element of guessing has been added. In rating these responses, you may wish to assign confidence rating weights to refleck the uncertainty which is often involved. Thus, if you are uncertain but feel that Inside Inter- pretation is more likely than Implication Re- flection, assign weights of "2" and "1" respec- tively to the two categories. On the other hand, if Reflection is the more likely category, then give it "2" and Interpretation a "1”. In general, when in doubt, Reflection should be slightly favored over Interpretation; however, there may be rare borderline cases which are impossible to judge--these should be given weights of "2" for both Reflection and Inter- pretation. Interpretations often carry markers or clues: These include signs of tentativeness ("It seems", "Perhaps", "I'm not sure, but, "Maybe," "It appears that," etc.) and technical or psychological labels ("inferiority complex," "Underdog," "Parent," "Snow-balling," etc.). However, these clues can be misleading, so they should only be used to alert you to the possibility that a response may be an Interpretation. In general, when you're in doubt about whether a response is an Interpretation or a Reflection, it's better to err on the side of Reflection. \_I . REASSURANCE Reassurance consists of the therapist responding positively E2_the client in some way. There are major and mihor fOrms of reassurance. A. MAJOR REASSURANCES ‘Major reassurances are positive about specific behaviors, qualities or outcomes of the client. 1. AGREEMENT REASSURANCES occur when the therapist agrees with the truth, feasibility or advisability of something the client has described or proposed. "That's really true." "Yes, that sounds like it might work." (N: Probably would be rated as General Advise- ment also.) Many agreements are small, short, one-word types-- "Yes", "OK", "Right", and sometimes "Uh-huh." However, when these introduce or are connected with a longer response, they should be disregarded. e.g., "Yeah, you're feeling really down today." (Do not code this response as Reassurance.) When short-agreements occur by themselves as the whole of a therapist's turn or as a separate rating unit, then they should be assigned a weight of "2". Finally, when "Uh-huh" is an agreement, then it should be given a "2" rating. 2. SUPPORT REASSURANCES describe the client's abilityto c0pe with a prohlem; in addition they sometimes predict a positive future outcome. 'I' "You can do it. (Rate also as General Advisement.) "It's going to turn out OK." (Rate also as Prediction Interpretation.) '3. PRAISE. The helper describes what the client has done or thought in positive or strongly 129 130 positive terms. Praise also sometimes points out a quality or strength in the client or a positive change the client has made. "It's a great step forward that you were able to stand up to your mother and tell her to get off your case." (Rate also as Interpretation.) "Bully for you!" "You know, you have a warm, engaging manner; I've) really enjoyed talking with you." ‘ (Rate also as Interpretation and Self-disclosure.) 4. MINIMIZERS are Reassurances in which the therapist tries directly to make the client feel better or more comfortable, by playing down or normalizing the client's concerns, problems, or self-criticisms. C: Sometimes-I feel like I'll never be able to think of a good example of a minimizing reassurance." " ” T: "It's really not as bad as you think." (Note: May be Interpretation also—- depending on context.) T: "Everyone has that feeling sometime." (Note: Information) T: "Don't worry-- you'll get over it." (Note: Assign a weight of "2" for Advisement and Interpretation also.) Also included are disagreements with self-criticisms. C: "Don't you think I'm a horrible person?" T: "No, I think you have many outstanding qualities." Note that all Minimizers are doubly coded as Reassurances and Disagreement. 5. PERMISSION REASSURANCES allow or give the client the OK to do something. C: "I shouldn't cry." T: "It's OK-- I don't mind." Note: Assign a weight of "2" for Self-disclosure and Process Advisement. 131 C: "I don't know what to say." T: "You don't have to say any thing." ("2” for Process Advisement) Note that these are usually also Process Advisements. B. MINOR REASSURANCES In addition to the five major forms of Reassurance, there are three minor types which do not address specific client behaviors or questions. '6. SYMPATHY is Reassurance in which the therapist commiserates or lodges a complaint similar to the client's. C: "Yeah, I just don't know what to do. I don't know how I'm going to pay the rent and I have all these bills and ..." T: "Yeah, it's really a hard situation to be in." T: "That's really difficult." N: Assign a weight of "2" to these. 7. CLAIMS OF UNDERSTANDING include brief statements of this—Eort: "I see", "I know just what you mean", "I gotcha." N: Assign these a weight of "2". 8. ACKNOWLEDGEMENTS (isolated). These include simple signs that the therapist is paying attention: "Uh-huh," "M-hm," "Mm,” "That's interesting." Note: When these occur in isolation (not connected to a longer response), they should be assigned a weight of "1". If they introduce a longer response, they should be dis- regarded. When acknowledgements are actually agreements (see above), then they should be given ratings of ”2". y__I_ . DISAGREEMENTS In Disagreement, the therapist differs, corrects, questions the assumptions of, or contradicts the client; there is usually an explicit or implicit disagreeing tone, although? it may be disguised. Also included are refusals of client requests for help, information, or agreement. Note that Disagreements are almost always coded as something in addition (often Advisement, Self-disclosure, Interpre- tation or Question). It is useful to identify 7 subtypes: 1. SIMPLE DISAGREEMENTS differ with what the client has said in some way. They contradict or deny the truth of something the client has said or the advisaEiIity of something the client proposes to do. Negation words, such as "No", ”not", "but", (also "Yes, but") frequently indicate simple disagreement (but there are exceptions). "That's probably not a good idea." (N: Also give a "2" General Advisement.) "Yes, but don't think there might be another reason for your behavior?" (N: Depending upon context, usually this would get "2” for Interpretation) "I really don't think that's going to work." ("2" for General Advisement) "I'd really rather you didn't smoke pot before coming to our session." (N: Rate also as ”2" Self-disclosure and "3" General Advisement) There is an exception to the negation-word role: C: ”I really don't want to work this summer." T: "No, you really don't." (Rate "3" for Reflection) DISCREPANCY DISAGREEMENTS point out contradictions or discrepancies in what the client has said or done.. They generally (but not always) have two parts, separated by a contradiction indicator (e.g., "but"). The discrepancy may be between past and present behavior or statements by the client, between verbal and nonverbal behavior, or between different"parts" (i.e., states, wishes, feelings) of the client. 132 Note: 133 "Last week you said you were ending the rela- tionship, but today you seem to want to work things out." (N: Summary Reflection also) ”How can you say you're relaxed when you can't sit still?" (N: Non-verbal Reflection Question) C: ."I really don't see the point of all this." T: "But you still keep coming back every week." Do not code a response as a Disagreementif the helper is Reflecting a contradiction stated by the client. 3. WARNINGS are both Disagreements and Advisements (Process or General). They attempt to get the client to not do something by warning him/her of negative consequences. "I think there'll be trouble if you do that." (N: "2" for General Advisement) ”But what if your boyfriend finds out you've been going out with someone else?" (N: ”3" for Question, "2" for General Advise- ment) "If you can't think of anything to say, then we may have to end our session early today." (N: "2” for Process Advisement) CRITICISM is a strong form of Disagreement in which the helper attacks, criticizes, scolds, blames, ridicules or puts down the client. Also included is sarcasm, which may rely in part on context and tone of voice. "I've got to admit, that was a pretty dumb thing to do." (N: Interpretation or Reassurance depending on Context) "You really shouldn't have let them walk all over you." (N: depending on context, may code a "2" for implied General Advisement that client should do better next time.) "Do you really think you should go on job interviews w1t out getting a haircut?" (N: "3" for General Advisement) C: "I've just invented a new kind of waterspout!" T: "I'm overwhelmed did you take your meds ‘ 97' . E§?8y“2" for Disagreement and Self-Disclosure; second part-—Closed Question "3") Note: 7. 134 QUESTIONING DISAGREEMENTS occur when the thera- pist asks a Question (often of the Closed type) or series of Questions about the client's assumptions or evidence, in order to throw doubt on them. C: "I'll never be able to get away from my parents." T: "Didn't you go to Europe without them last summer?" C: "Well, yes ---" T: "Didn't you get a job so you didn't have to borrow from them again?" C: "Well, yes---" T: "Well, then, what makes you think you can't survive without them?" (Note: Questioning Disagreements should be assigned a weight of "2". CORRECTION DISAGREEMENTS occur when the therapist differs with the client, and offers new substitute information in place of something the client has said or may believe. This includes cases in which the therapist disagrees with the client as the therapist has interpreted the client C: "I would work and turn the paycheck over to the family because they were having emotional and financial problems...you know...a lot of family stress at this time." T: ”So your family was under a lot of stress but it sounds like you were under a lot of stress yourself.” (N: Rate Reflection "3.") C: "And I don't know where I get that from." T: "I don't know that you got to get it from any place, it's just nice to be pursued by the opposite sex." (N: Reassurance: "2"; interpretation: "2.") The maximum rating for correction disagreements is "2". HIDDEN DISAGREEMENTS signal Disagreement without actually stating it. The therapist implicitly differs with the client by usually (but not always) refusing a client's request for help, information, agreement or reassurance. Hint: Look for client questions immediately preceding the therapist response. 135 "Are you married?‘ "Why do you ask?" "Tell me what you think I should do." "I wish I could tell you." HOHO Note: Maximum rating for hidden disagreements is "2". VII. SELF—DISCLOSURE Self-disclosures are responses in which the therapist deliberately refers to him/herself in order to reveal some- thing significant about him-herself—-a personal experience (past or present), intention, goalj'or limitation. Self— disclosures do not always have to use the pronouns "I" or "me . l! 1. There are two major subtypes: GENERAL SELF-DISCLOSURES reveal some past or present personal expefience or quality of the therapist which is focused outside the helping session. "When I was a kid, I felt shy and lonely also.” "I'm an oldest child also." T: "I had the same problem when I was in college, too." C: "Do you ever get tired of listening to people?” T: "Not tired, but you sometimes get the feeling you need time to talk things over with someone yourself." PROCESS SELF-DISCLOSURES reveal something about what is going on with the therapist within the helping session--a personal reaction or association to what the client has been saying, an intention goal or promise to do something, or a limitation of personal expertise or knowledge. "When you say that, it makes me feel like crying too.” "I'm feeling frustrated because you don't seem to like anything I try to do, but you won't tell me what you want either." (N: Code "3" for Disagreement also.) C: "Can you help me?" T: "I can't promise to help you, but I do promise to try to help you." (Code "1" for Disagreement also) or "We'll do what we can to help you." "I'm having trouble following you and I feel lost." "When I asked you all those questions, I 136 137 wasn't trying to gather information, I was trying to get you to explore those issues." (Probably rate "2" for disagreement) C: "I'm feeling angry." T: "I think we're both feeling angry about this." Note: 29 nOt code stylistic self-disclosures in which the therapist uses self-reference to soften (attenuate) or introduce some other type of behavior (e.g., Interpretation, Disagreement, Question). "I think ---", "It seems to me ---", "My opinion is ---", "I wonder---", "If I were you, I'd---", "I mean---", "To me it's--—” Stylistic self-disclosures include "I" type Open Questions and, in general, self-disclosures of the therapist's o inions about the client (i.e., Interpretations) or what the client should do (i.e., Advisements). VIII. INFORMATION Information responses attempt to give orienting information to the client. They aim to instruct the client by giving information not specifically about the client. These responses involve new information about the helping process, third parties, or people in general. 1. PROCEDURAL INFORMATION involves the helper giving simple instructions about helping situations, including locations, scheduling, time limits, administrative procedures, fees, and referrals. "This relaxation tape lasts about 20 minutes." "This psychology clinic is open from 9:00 to 5:00, and is on the first floor of University Hall." "We have a sliding fee schedule." "Your case will be presented in staff meeting on Friday: someone will call you on Monday." THIRD PARTY INFORMATION gives new information about specific people other than the client or therapist. These responses may often go beyond simple information to make interpretations about other specific people. "Your mother valued control because she was afraid of being crazy herself." C: "What do you know about Dr. Hill?" T: "I've heard that she is a very good therapist. "It's my opinion that your family is using you as a scapegoat." (N: This should also be scored as a "3" for Interpretation or the client's role in the family.) . 'GENERAL INFORMATION involves the therapist draw- ihg on his/her expertise to tell the client about people in general, general psychological princi- ples, or treatment rationales. "Relaxation can be seen as a skill which you can learn like any other." "Desensitization treatment has been found to be useful for helping people with spider 138 F' 139 phobias to overcome their fears. (N: This also gets a ”2" for Informational Advisement.) "I feel lonely since I got to college." "Many people feel lonely their first year away at college." (N: Rate "3" for minimizing reassurance.) IX. OTHER This category is for types of therapist behavior which don't fall anywhere else, but which are still recog- nizable data. There are three major subtypes of Other: 1. SOCIAL TALK includes social rituals, especially greetings,introductions, endings, and jokes. "Hello, my name is Mr. Helper." "The microphone goes on like this." "Goodbye and good luck." "Have a nice day." 2. SELF-TALK occurs when the therapist thinks outloud or mutters to her/himself (usually self-reference is present). ”'Foos around?’ Am I so old? "Foos around?'" ”Let me see, how should I ask this?" These are not to be coded as Questions. 3. OTHER-OTHER QR WASTEBASKET. When a response doesn't fit anywhere else, it should be relegated to this subcategory. (That means that this category is mutually exclusive of all others.) I " "Yeah, gee." 'H ”Woweee. 140 X: ADDITIONAL NOTES l. REPETITIONS AND REPAIRS. Sometimes a therapist will repeat or correct What he or she said earlier. In these cases, the second occurrence should be coded the same as the first. For example, T: "How long has this been going on?" (Closed Question) C: "This?"w T: "I mean, how long have you been having problems communicating with your children?" (Closed Question) This also applies to advisements and interpretations. 2. EXCLUDED DATA. Some types of therapist behavior should be either ignored altogether or treated as missing data and excluded from the process of coding and estimating reliability. There are three major categories of these: (a) Fragments are false starts in which the therapist does not finish her/his thought; usually this means that the predicate of the sentence is missing. For example, "Well, I--" The therapist may stop on his/her own or may be interrupted by the client. Note that Collaborative Reflections and Fill- in Questions both do involve complete thoughts or acts and so should be coded. (b) Inaudible 9}; incomprehensible responses should also be excluded from consideration. e.g. T: I! )H. C: "What?" (c) Other noises should also be excluded. These include: laughter, coughing, other people talking, environmental noises and so on. Note: If you are asked to rate such responses, you should indicate that no response modes are present--i.e., by giving the response "0" for all categories. 141 APPENDIX B CLIENT VERBAL RESPONSE CATEGORY SYSTEM Copyright (c) C. E. Hill, C. Greenwald, K. G. Reed, D. Charles, M. K. O'Farrell, and J. A. Carter, 1981 (Adapted for use in this study) 142 APPENDIX B CLIENT VERBAL RESPONSE CATEGORY SYSTEM Introduction to Client System Much of the literature has focused on counselor verbal behavior to the exclusion of client verbal behavior. This undoubtedly reflects the emphasis on training counselors in techniques. Unfortunately, this overemphasis on counselor behavior has limited the awareness of the clients' response. It seems very important to know what clients do in the session for diagnostic purposes, e.g., are they able to express their feelings or be insightful? The client's behavior determines how the counselor will respond. Further, a comprehensive client category system is important to study how the client responds to counselor interventions, e.g., how does the client respond when the counselor uses reflections vs. interpretations? This system was based on Snyder's (1945) original category system for measuring client behavior. Development of the system followed the same procedures as those for the counselor system, including application of the system to several transcripts and validation of the defini- tions and examples by five expert psychologists. The system has undergone four revisions before coming to this final version with nine categories. The system consists of nine categories: Client Categories 1. Simple Responses Requests Description Experiencing Exploration of Counselor-Client Relationship Insight \IO‘U'IJ-‘wN o. 0.0 Discussion of Plans 143 144 8. Silence 9. Other The complete system, including definitions and examples, is reprinted below. Client Verbal Response Category System 1. Simple Responses: A short and limited phrase mi (typically one or two wordS). Usually of three types: a) Indicates agreement, acknowledgement, understanding, or approval of what the counselor has said; b) indicates disagreement or disapproval with what the counselor has said; c) responds briefly to a counselor's question with specific information or facts. (Note: Just because the counselor asks a question, do not automatically put the client's response here. In fact, tend to put it in another category unless it is just a very simple response.) Generally, responses in this category do not indicate feelings, description, or exploration of the problem. (Note: Each of the three subtypes should be rated as separate categories in this study.) Examples: CO: "As I said before, you seem angry." CL: "You're right." (Agreement) CO: "You were really prepared for that test." CL: "MmHm." (Agreement) CO: "You would like to be more positive." CL: "1 don't really think that's a problem." (Disagreement) CO: "I think we should talk about these issues." CL: "No, I don't want to." (Disagreement) CO: "What grade did you get?" CL: ”C." (Simple Information) CO: CL: CO: CL: CO: CL: 145 "Did you really want to go there?" "No, not really." (Simple Information) ”Do you want to come at a different time?" ”Let me think about it." (Simple Info.) "Why do you think you did that?" "I'm not sure." (Simple Information) 2. Requests: Attempts to obtain information or advice or to place the burden or responsibility for solu- tion of the problem on the counselor. Examples: CO: CL: CO: CL: CO: CL: CO: CL: CO: CL: "You're not sure what to do." "What do you think I ought to do?" ”It sounds like you would be interested in being a nursery school teacher." "Where would I find out about the necessary qualifications for that?" "So you've given up disciplining the children?” "How do you think I should handle the children?" ”You've decided to finish the semester then.” "Do you think I should drop out?" (Pause = 10 sec.) "Where do you want me to begin?" 3. Description: Discusses history, events, or incidents related to the problem in a story-telling or narrative style. The person seems more interested in letting you know what happened rather than in their affective responses, understanding, or resolving the problem. Examples: CO: CL: ”What would you like to talk about today?" "My mother pulled a really dirty trick on me of telling my sister something I told her in complete confidence." 146 CO: "You look sad right now." CL: "My boyfriend stood me up over the weekend and didn't even call to apologize." CO: "So what's been going on?" CL: "My husband and I had an awful fight last night where he called me every name in the book." CO: "Describe more aboutyour problem with men." CL: ”Like yesterday when I saw this really attractive man in the elevator, he didn't even look at me or notice that I existed." 4. Experiencing; Affectively explores feelings, behaviors, or reactions about self or problems, but does not convey an understanding of causality. May indicate a growing awareness of behaviors or problems without necessarily understanding why they have occurred. Does pep refer to feelings toward counselor/counseling situation. (Note: Sometimes listening to the audio-tape is helpful to differentiate this category from description.) Examples: CO: "You sound very angry right now." CL: "All I could do was withdraw and feel sad, but maybe angry too." CO: "It sounds like you're afraid to stand up to your mother." CL: "But I wonder why I am since other people stand up to her okay." CO: "You've gotten very quiet." CL: "I feel blocked right now and am not sure what to say." CO: "You haven't been able to do what you wanted." CL: "I seem to lack self-confidence right now." CO: "You seem more cheerful today." CL: "I'm feeling better and stronger and think I'll be able to cope." 147 CO: "You have a frown on your face." CL: "I'm angry at myself because I should have known better." 5. Exploration of Client-Counselor Relationship: Indicates feelings, reactions, attitudes, ortbehaviors related to the counselor or the counseling situation. Does pp; refer to feelings which are not directed towards the counselor. Examples: CO: "How did you feel about my not remembering your name when I saw you in the hall?" CL: "I felt hurt that I was not important to you." CO: ”Could you tell me how you're feeling right now?" CL: "I'm scared that if I tell you, you'll get angry." CO: ”I don't understand why you're so quiet today." CL: "You've been talking so much, I could- n't get a word in edgewise." CO: ”I feel unsure about what we've done today." CL: ”This has been really helpful to me." CO: "Last week you were really upset with me." CL: "I came in today to tell you that I'm not getting any place here and want to quit." 6. Insight: Indicates that a client understands or is able to see themes, patterns, or causal relationships in his/her behavior or personality, or in another's behavior or personality. Often has an "aha" quality. Insight statements usually have an appropriate internalization quality, i.e., the client takes the appropriate responsi- bility rather than assuming too much or blaming the other person or using "shoulds" imposed from outside rather than inside. Statements explaining the "why" of behavior should indicate a logical and reasoned explanation rather than a rationalization. (Note: This may be hard to determine; give the client the benefit of the doubt that he/she is not rationalizing unless it is an obvious distortion.) Examples: CO: CL: CO: CL: CO: CL: CO: CL: CO: CL: 148 "You do seem to need a lot of attention." "Maybe having everybody waiting on me made me so spoiled that I can't get along without that." "Why do you get so hostile to your mother?" ”I just realized that I think it's because I didn't feel like she took care of me very well." "You seem to have trouble organizing your time.” "I think I waste a lot of time and don't organize well because I'm afraid of having free time and not knowing how or who to spend it with." "What did you do to make your parents so angry?" ”I used to think it was me, but they were so angry at each other that they didn't know how to express it and took it out on me instead.” "What do you think is happening when you yell?" "I think really I'm afraid that he's going to leave me." 7. Discussion of Plans: Refers to action-oriented plans, decisions,’future goals, and possible outcomes of plans. The client seems to have a problem-solving attitude here. Discussion of past plans are BEE included here. Should be actual plans rather than hypothetical ruminations about the various possibilities open to the client in the future (these would fit under description). Examples: CO: CL: CO: "You've changed a lot as a result of counseling." "I really feel like the next time the boss asks me to work overtime with- out pay, I will be more assertive and say 'no.'" "You need to be thinking about earning your own money soon." 149 CL: I'm going to start applying for .jobs, even though I probably won't get the ideal one immediately." CO: "What could you do about feeling over- whelmed by your commitments?" CL: "I've decided to discontinue one of my projects and to cut some hours at my part-time job." CO: ”When are you planning on moving to an apartment?” CL: "I've decided to look into a group house rather than living alone in an apartment." CO: ”I think it would be worth talking to him." CL: "I'll go home and tell him how frightened I am." 8. Silence: A pause of 5 seconds (4 seconds is close enough), is considered the client's pause if it occurs between the counselor's statement and a client's statement, within the counselor's statement, or immediately after a client's simple response. (Note: Silences are not rated in this study.) Examples: CO: "You look angry.” CL: (”No.”) Pause = 5 sec. CO: ”What are you thinking?" CL: Pause = 4 sec. ("I'm not sure I can articulate it ") CO: "What do you want to do?" CL: Pause = 5 sec. CO: "Would you like to come in again?" 9. Other: Statements which are unrelated to the client's proEIem, such as small talk or salutations, comments about weather or events, or any statements which do not seem to fit into other categories due to difficulties in trans- cription, comprehensibility, or incompleteness. Examples: "CO: "Hello." "CL: "It's really beautiful outside today." CO: CL: CO: CL: 150 "Bye . I! ”See you next week." "You're looking cheerful." "The Redskins game was terrific." APPENDIX C SYMPTOM DISTRESS CHECKLIST CLIENT FORM (SCL—9OR) Copyright (c) Leonard R. Derogatis, Ph.D., 1975 151 INSTRUCTIONS: CATEGORIES: EXAMPLE: 1. Headaches APPENDIX C SYMPTOM DISTRESS CHECKLIST CLIENT FORM (SCL-90R) Below is a list of problems and complaints that people sometimes have. Please read each one carefully. After you have done so please circle one of the numbers to the right that best describes how much that problem has bothered or distressed you during the past couple weeks including today. Circle only one number for each problem and do not skip any items. Please read the example before beginning. 0 — Not at all 1 — A little bit 2 — Moderately 3 - Quite a bit 4 - Extremely How much were you bothered by l. Backaches.. By circling #1, this person answered that he/she was a little bit bothered by backaches. 2. Nervousness or shakiness inside 0 1 2 3 4 3. Unwanted thoughts, words, or ideas that won't leave your mind 4. Faintness or dizziness 0 l 2 3 4 5. Loss of sexual interest or pleasure 6. Feeling critical of others 0 l 2 3 4 152 153 CATEGORIES: 0 - Not at all 1 - A little bit 2 — Moderately 3 - Quite a bit 4 — Extremely 7. The idea that someone else can control your thoughts 8. Feeling others are to blame for most of your troubles 9. Trouble remembering things 10. Worried about sloppiness or carelessness 11. Feeling easily annoyed or irritated 12. Pains in heart or chest 13. Feeling afraid in open spaces or on the streets 14. Feeling low in energy or slowed down 15. Thoughts of ending your life 16. Hearing voices that other people do not hear l7. Trembling 18. Feeling that most people cannot be trusted 19. Poor appetite 20. Crying easily 21. Feeling shy or uneasy with the opposite sex 22. Feeling of being trapped or caught 23. Suddenly scared for no reason 154 CATEGORIES: O - Not at all 1 - A little bit 2 - Moderately 3 - Quite a bit 4 - Extremely 2 4. Temper outbursts that you could not control 25 . Feeling afraid to go out of your house alone 26 . Blaming yourself for things 2 7 . Pains in lower back 28 Feeling blocked in getting things done 2 9 Feeling lonely 30 Feeling blue 3 J. Worrying too much 32 - Feeling no interest in things 33 - Feeling fearful 34 - Your feelings being easily hurt 35 . Other people being aware of your private thoughts 36 . Feeling others do not under- stand you or are unsympathetic 37 . Feeling that people are unfriendly or dislike you 38 . Having to do things very slowly to insure correctness 39 . Heart pounding or racing 40 - Nausea or upset stomach 41 . Feeling inferior to others 0 0000000 0000 H l—‘l—‘t—‘t—‘i—‘t—‘H t—‘D—‘t—IH N NNNNNNN NNNN CO wwwwwww wwww b b-L‘b-L‘b-L‘b b-L‘J-‘b 155 (:34ATEGORIES: 0 - Not at all 1 - A little bit 2 - Moderately 3 - Quite a bit 4 - Extremely ¢Qp12. Soreness of your muscles 43—13. Feeling that you are watched or talked about by others 43.43. Trouble falling asleep £¥.ES. Having to check and double- check what to do éytES. Difficulty making decisions 5&77'. Feeling afraid to travel on buses, subways or trains 48 . Trouble getting your breath 49 . Hot or cold spells 50. Having to avoid certain things, places, or activities because they frighten you 5 l . Your mind going blank 52. Numbness or tingling in parts of your body 53 . A lump in your throat 54. Feeling hopeless about the future 55 . Trouble concentrating 56 . Feeling weak in parts of your body 57 - Feeling tense or keyed up 58 . Heavy feelings in your arms or legs 156 CATEGORIES: O - Not at all 1 - A little bit 2 - Moderately 3 - Quite a bit 4 - Extremely 59. Thoughts of death or dying 60. Overeating 61. Feeling uneasy when people are watching or talking about you 62. Having thoughts that are not your own 63. Having urges to beat, injure, or harm someone 64. Awakening in the early morning (55. Having ideas or beliefs that others do not share (56. Sleep that is restless or disturbed 6 7 Having urges to break or smash things 58 Having ideas or beliefs that others do not share 69 . Feeling very self-conscious with others 7'C) . Feeling uneasy in crowds such as shopping or at a movie 7 l . Feeling everything is an effort .7'33 . Spells of terror or panic 77§33 . Feeling uncomfortable about eating or drinking in public :7'<¥-. Getting into frequent arguments 157 CATEGORIES: 0 — Not at all 1 - A little bit 2 - Moderately 3 - Quite a bit 4 - Extremely _75. Feeling nervous when you are left alone 76. Others not giving you proper credit for your achievements 77. Feeling lonely even when you are with people 78. Feeling so restless you couldn t sit still 79. Feeling of worthlessness 80. Feeling that familiar things are strange or unreal 81. Shouting or throwing things 82. Feeling afraid you will faint in public £33. Feeling that people will take advantage of you if you let them 84. Having thoughts about sex that bother you a lot 85 . The idea that you should be punished for your sins 86 . Feeling pushed to get things done 8 7 . The idea that something serious is wrong with your body 88 . Never feeling close to another person 89 Feelings of guilt 158 CATEGORIES: 0 - Not at all 1 — A little bit 2 - Moderately 3 - Quite a bit 4 - Extremely 90. The idea that something is wrong with your mind APPENDIX D SYMPTOM DISTRESS CHECKLIST CLINICIAN FORM (SCL-9OA) Copyright (c) Leonard R. Derogatis, Ph.D. and Nick Mellisaratos, 1976 p-a U1 \0 APPENDIX D SYMPTOM DISTRESS CHECKLIST CLINICIAN FORM (SCL-9OA) HOPKINS PSYCHIATRIC RATINGS CATEGORIES: 0 — None 1 — Slight 2 - Mild 3 - Moderate 4 - Marked 5 - Severe 6 — Extreme 1. Somatization 0 l 2. Obsessive-Compulsive O l 3. Interpersonal Sensitivity 0 1 4. Depression 0 l 5. Anxiety 0 1 6. Hostility 0 l 7. Phobic Anxiety 0 1 8. Paranoid Ideation 0 l 9. Global Pathology Index 0 1 160 NNNNNNNNN wwwwwwwww b-L‘bbb-Dbbb U'IU1U1U'IU'IU'IU1LIIU1 O‘C‘O‘O‘O‘O‘O‘O‘O‘ APPENDIX E POSTTHERAPY CLIENT QUESTIONNAIRE 161 APPENDIX E POSTTHERAPY CLIENT QUESTIONNAIRE For each item choose the anSwer which you feel best describes your therapy experience. 1. How much in need of further therapy do you feel now? No need at all Slight need Could use more Considerable need Very great need I 2. What led to the termination of your therapy? My decision My therapist's decision Mutual agreement External factors *3. How much have you benefitted from your therapy? A great deal A fair amount To some extent Very little Not at all *4. Everything considered, how satisfied are you with the results 0 your psychotherapy experience? Extremely dissatisfied Moderately dissatisfied Fairly dissatisfied Fairly satisfied oderately satisfied Highly satisfied Extremely satisfied *Questions used in this study. 162 163 What impression did you have of your therapist's level of experience? Extremely inexperienced Rather inexperienced Somewhat experienced Fairly experienced Highly experienced Exceptionally experienced How well did you feel you were getting along before therapy? Very well Fairly well Neither well nor poorly Fairly poorly Very poorly Extremely poorly How long before entering therapy did you feel in need of professional help? 1 Less than 1 year 1 — 2 years 3 - 4 years 5 — 10 years 11 — 15 years 16 - 20 years How severely disturbed did you consider yourself at the beginning of your therapy? Extremely disturbed Very much disturbed Moderately disturbed Somewhat disturbed Very slightly disturbed A tremendous amount A great deal A fair amount Very little None at all 164 10. How great was the internal "pressure" to do something about these problems when you entered psychotherapy? Extremely great Very great Fairly great Relatively small Very small Extremely small *11. How much do you feel you have changed as a result of psychotherapy? A great deal A fair amount Somewhat Very little Not at all 12. How much of this change do you feel has been apparent (a) People closest to you (husband, wife, etc.) A great deal A fair amount Somewhat Very little Not at all (b) Close friends A great deal A fair amount Somewhat Very little Not at all (c) Co-workers, acquaintances, etc A great deal A fair amount Somewhat Very little Not at all 13. On the whole, how well do you feel you are getting along now? Extremely well Very well Fairly well ' Neither well nor poorly Fairly poorly Very poorly Extremely poorly *Questions used in this study. 14. *15. 16. 17. 165 How adequately do you feel you are dealing with any present problems? Very adequately Fairly adequately Neither adequately nor inadequately Somewhat inadequately Very inadequately To what extent have your complaints or symptoms that brought you to therapy changed as a result of treatment? Completely disappeared Very greatly improved ___Considerably improved Somewhat improved -Not at all improved Got worse How soon after entering therapy did you feel any marked change? weeks of therapy (approximately) How strongly would you recommend psychotherapy to a close friend with emotional problems? , Would strongly recommend it Would mildly recommend it Would recommend it but with some reservations —”__FWould not recommend it _____Would advise against it Please indicate to what extent each of the following statements describes your therapy experience. Disregard that at one point or another in therapy you may have felt differently. l8. l9. 1 - Strongly agree 3 - Mildly agree 5 - Undecided 7 - Mildly disagree 9 - Strongly disagree The following questions were rated on the above scale. My therapy was an intensely emotional experience. My therapy was often a rather painful experience. RQuestions used in this study 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 166 - Strongly agree - Mildly agree Undecided - Mildly disagree - Strongly disagree \OVMUJH U I remember very little about the details of my psychotherapeutic work. My therapist almost never used technical terms. On the whole I experienced very little feeling in the course of therapy. There were times when I experienced intense anger toward my therapist. I feel the therapist was rather active most of the time. I am convinced that the therapist respected me as a person. I feel the therapist was genuinely interested in helping me. I often felt I was "just another patient." The therapist was always keenly attentive to what I had to say. The therapist often used very abstract language. He very rarely engaged in small talk. The therapist tended to be rather stiff and formal. The therapist's manner was quite natural and unstudied. I feel that he often didn't understand my feelings. I feel he was extremely passive. His general attitude was rather cold and distant. I often had the feeling that he talked too much. I was never sure whether the therapist thought I was a worthwhile person. I had a feeling of absolute trust in the therapist's integrity as a person. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 167 - Strongly agree Mildly agree Undecided - Mildly disagree - Strongly disagree \ONU'IUJH I I felt there usually was a good deal of warmth in the way he talked to me. The tone of his statements tended to be rather cold. The tone of his statements tended to be rather neutral. I was never given any instructions or advice on how to conduct my life. The therapist often talked about pschoanalytic theory in my sessions. A major emphasis in treatment was upon my attitudes and feelings about the therapist. A major emphasis in treatment was upon my relation- ships with people in my current life. A major emphasis in treatment was upon childhood experiences. A major emphasis in treatment was upon gestures, silences, shifts in my tone of voice and bodily movements. I was almost never given any reassurances by the therapists. My therapist showed very little interest in my dreams and fantasies. I usually felt I was fully accepted by the therapist. I never had the slightest doubt about the therapist's interest in helping me. I was often uncertain about the therapist's real feelings toward me. The therapist's manner of speaking seemed rather formal. 54. 55. 168 Strongly agree Mildly agree Undecided Mildly disagree Strongly disagree \DVLHUDH I feel the emotional experience of therapy was much more important in producing change than intellectual understanding of my problems. My therapist stressed intellectual understanding as much as emotional experiencing. APPENDIX F POSTTHERAPY THERAPIST QUESTIONNAIRE 169 APPENDIX F POSTTHERAPY THERAPIST QUESTIONNAIRE HOPKINS PSYCIATRIC RATINGS Categories: 0 - None 1 - Slight 2 - Mild 3 — Moderate 4 - Marked 5 - Severe 6 - Extreme 1. Somatization 0 l 2. Obsessive-Compulsive 0 l 3. Interpersonal Sensitivity 0 1 4. Depression 0 l 5. Anxiety 0 l 6. Hostility 0 l 7. Phobic Anxiety 0 l 8. Paranoid Ideation 0 1 9. Psychoticism 0 l 10 . Global Pathology Index 0 l 170 NNNNNNNNNN wwwwwwwwww .p -b a» c~ 4> -b D- ¢~ 4> -b U1U'IUIU‘IU1U1U1UTU'IU'I O‘O\O\O\O\O\O\O\C\O\ 171 Please rate each of the following items, comparing the client with other clients whom you see in psychotherapy using the following scale: 1 - Very little 3 - Some 5 — Moderate 7 - Fairly great 9 — Very great 11. Defensiveness l 3 5 7 12. Anxiety 1 3 5 7 l3. Ego Strength 1 3 5 7 14. Degree of disturbance l 3 5 7 15. Capacity of insight 1 3 5 7 16. Overall adjustment 1 3 5 7 17. Personal like for patient 1 3 5 7 18. Motivation for therapy 1 3 5 7 19. Improvement expected (Prognosis) l 3 5 7 20. Degree to which counter- transference was a problem in therapy 1 3 5 7 21. Degree to which you usually enjoy working with this kind of patient in psychotherapy 1 3 5 7 *22. Degree of symptomatic improvement 1 3 5 7 23. Degree of change in basic personality structure 1 3 5 7 24. Degree to which you felt warmly toward the patient 1 3 5 7 *Questions used in this study. \OKDKDKOKDKOOQ 172 l - very little 3 - Some 5 - Moderate 7 - Fairly great 9 — Very great 25. How much of an "emotional investment" did you have in this patient? 1 3 5 7 9 26. Degree to which you think the patient felt warmly toward you. 1 3 5 7 9 *27. Overall success of therapy 1 3 5 7 9 28. How would you characterize your working relationship with this patient? 1 3 5 7 9 1 Extremely poor - Fairly poor Neither good nor poor - Fairly good - Extremely good oume l *29. How satisfied do you think the patient was with the results of his therapy? 1 3 5 7 9 - Extremely dissatisfied - Fairly dissatisfied Neither satisfied nor dissatisfied - Fairly satisfied - Extremely satisfied \O\J £11901—l l 30. How would you characterize the form of psychotherapy you conducted with this patient? 1 3 5 7 9 l 3 5 7 9 Largely supportive Intensive analytical *Questions used in this study. 173 31. Do you recall any strikingly pleasant experiences that you had during the therapy sessions with this patient? If yes, please mark the number that best indicates the degree of pleasantness. Otherwise mark "0" for No. l 2 3 4 5 6 7 8 9 l 2 3 4 5 6 7 8 9 Mildly pleasant Extremely pleasant 32. Do you recall any strikingly unpleasant experiences you had with this patient? If yes, please mark the number that best indicates the de- gree of unpleasantness. Otherwise mark ”0" for No. l 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 Mildly pleasant Extremely pleasant 33. Overall, how would you characterize your exper- iences with this patient? 1 3 4 5 6 7 8 9 l 2 3 4 5 6 7 8 9 N 1— Unpleasant Pleasant APPENDIX G STANDARD (T—SCORE) NORMS FOR MICHIGAN STATE PSYCHOLOGICAL CLINIC CLINICIAN RATINGS OF MALE AND FEMALE OUTPATIENTS ON THE SCL-90A'S SYMPTOM DIMENSIONS AND GLOBAL PATHOLOGY INDEX. From Filak, J. (1982). Congruence of perception on client symptoms and therapy outcome. Unpublished thesis, Department of Psychology Michigan State University. 174 175 N.Hn w.mo q.cm mq o.H¢ N.¢m w.om Haw H.moH 3.8 Ham 4:2 #8 #3 #3 Ham .MOH .mm .mo .mm mq Hmm o m «.mm m N H m.mm m.mw N.wm On m.Ho o.mm m.mq nose .om .Nw .mm .mo .mm NQOQON .mq Hq mom Nmmoo