Which of the following terms indicates a caring relationship between a client and therapist?

J Psychother Pract Res. 1998 Winter; 7(1): 56–64.

Abstract

Although disparate views have been published, the theory underlying multimodal therapy suggests that therapist–client similarity would be most advantageous for treatment outcome and client satisfaction. To explore this question, 19 different therapist–client pairs were followed over 12 sessions of psychotherapy. Clients were evaluated with the Brief Symptom Inventory (BSI) after sessions 1 and 12 to determine psychotherapy outcome. Similarity was determined by computing D″2 statistics on therapists' and clients' responses to the Structural Profile Inventory (SPI). Similarity on the SPI predicted psychotherapy outcome, showing a statistically significant relationship with the Global Severity Index of the BSI.

In clinical psychology's quest to fulfill G. L. Paul's1 charge to psychotherapy researchers, “What therapy, by whom, is most effective for this individual with that specific problem and under which circumstances?” (p. 111; italics in original), research has shown that it is not only important to study each of these factors individually, but that it is also important to combine them. The study of therapist–client similarity has thus arisen. Out of this tradition of research, two opposing schools of thought have developed: one suggesting that therapist–client similarity results in optimized outcomes, and one suggesting that dissimilarity optimizes treatment outcomes.

The position favoring similarity appears to have evolved from observations that therapist and client demographic and personality characteristics such as gender, race, personality, and mental health have a “profound impact” on psychotherapeutic process and outcome.2 Some research in this area has gone so far as to suggest that most of the variance in outcome is a result of therapist and client variables and that little is actually a result of the specific techniques used.3, 4

It has also been suggested that client–therapist similarity aids in the genesis and maintenance of rapport.5 Psychotherapy is an interpersonal sharing and communication experience. Therefore, the more similar the client–therapist dyad, the greater the likelihood of the communication being clear and readily understood.6 Dormarr et al.7 demonstrated that the clearer and more consistent the communication between the therapist and client, the more positive the treatment outcome. Cummings et al.8 found that when therapists and clients agreed in their recollections of which session events were important, psychotherapy sessions were generally rated as more effective by both clients and therapists.

Strong support for this position comes from studies of length of psychotherapy as a function of client–therapist similarity. Mendelson and Geller9 studied therapists and clients in a college counseling center. Similarity/dissimilarity was determined on the basis of Myers-Briggs Type Inventory (MBTI) profiles. Dissimilarity was associated with increased dropout rates and increased premature termination rates. Purportedly, clients were unable to build rapport with highly dissimilar therapists and left therapy because of dissatisfaction.

Much of the support favoring dissimilarity comes from the perspective that psychotherapy is an educational experience. There is a large body of literature10,11 suggesting that the “active ingredient” of psychotherapy is the set of new or compensatory skills acquired during the process. For education to be effective, the “student” must be presented with some information that is new or that is presented in a different way. In psychotherapy, clients must be able to learn something novel and different from their therapists that they would not ordinarily learn on their own. Therefore, therapists who are too similar to their clients will be unable to present a different perspective or any new learning.12 Even studies that have shown the importance of client–therapist similarity have demonstrated that extreme similarity appears to be a deterrent to successful psychotherapeutic outcome for precisely this reason.7

The study of the effects of therapist–client matching of personality variables is not new to the science of psychotherapy. In the early 1970s a flurry of research activity was generated on the topic of the A-B dichotomy.13–16 According to this theory, certain therapist personality variables, differentiable through the administration of the Strong Vocational Interest Blank, could predict outcome efficacy with neurotic versus psychotic patients. However, this theory has not been validated empirically.17

More recently, Beutler and Clarkin18 developed a system for systematically targeting therapeutic interventions for specific clients with specific problems in specific circumstances. This integrative model is based on four primary classes of events and variables, which therapists can consider and use to theoretically choose optimal psychotherapeutic techniques to maximize treatment outcome. These four primary classes are Patient Predisposing Variables (diagnosis, personal characteristics, environments/circumstances), Treatment Context (setting, mode or format, frequency and duration), Relationship Variables (personal compatibility matching, enhancing of the therapeutic alliance), and Strategies and Techniques (focal targets of change, level at which goals of treatment are mediated, way of conducting the actual therapeutic work).

It is the third of these classes, Relationship Variables, that is most appropriate to this report. Among the dimensions of compatibility discussed by Beutler and Clarkin are demographics, interpersonal response patterns, personal beliefs, and attributions. (See their chapter 9 for a comprehensive discussion of therapist–client personal and personality matching in the optimization of psychotherapy outcome.18)

Multimodal therapy19 explains human functioning in terms of seven independent yet interactive dimensions, referred to by the acronym BASIC-I.D. These seven spheres of functioning—Behaviors, Affects, Sensory, Imagery, Cognitions, Interpersonal, and “Drugs”/biological factors—either singly or in combination can explain fully the realm of human experience and functioning.19

The modality of Behaviors can best be described as one's orientation to action. People who score high on the Behaviors modality scale are generally described as active, energetic, and busy. They are often goal-oriented and often choose to act on a problem rather than studying it in depth first. People who score high on the Affects modality consider themselves emotional. They feel things deeply and rely on their emotions and intuitions. People who score high in the Sensory modality are very tuned in to their physical sensations. They are keenly aware of smells, tastes, sights, kinesthetics, and sounds, similar to the conceptualization of the strongly right brain–dominant individual. People who score high on the Imagery modality are good at thinking in pictures. They may be more likely to fantasize or daydream and can often think three-dimensionally. People who score high in the Cognitions modality consider themselves logical, rational, and contemplative. People who score high in the Interpersonal modality derive energy from interpersonal relationships. These are “people persons” who like to socialize, mingle, and be in groups. People who score high in the Biochemical factors are health conscious. They avoid unhealthy habits and take care of their bodies. They do not resort to substance use to cope. People who are having psychological problems will experience them across all modalities. Consequently, for optimal treatment outcome, therapy must focus on the significant manifestations in all seven modalities.

In Multimodal therapy, clients are thoroughly assessed in all seven areas of functioning. All of their strengths and clinical symptoms in all modalities are carefully cataloged. The Multimodal Treatment Plan is developed by matching specific, empirically documented psychotherapeutic techniques to each of the client's target clinical symptoms. Multimodal therapy agrees with other disciplines and psychological views20,21 in recognizing that it is the techniques that are of prime importance in psychotherapy.19

Modality scores are important in the study of behavior. They determine an individual's functioning preferences. According to the theory and clinical observation of multimodal theory,19,20 one's “dominant modality” (having the highest score on the Structural Profile Inventory [SPI]) will be the sphere of functioning in which one will be most likely to react, especially in times of stress. The implications of this finding for marital and other interpersonal relationships are that when individuals share dominant modalities, their communication will most likely be clearer and a more productive interpersonal relationship will result.22–24

Dominant modalities can be determined through the application of several different assessments. The Structural Profile was originally a verbally administered tool25 that consisted of describing the seven modalities and requesting the client's self-rating for each. Straightforward in its approach, the Structural Profile is a quick and easy way to obtain a general and global picture of a client's modality functioning. However, to gain further insight into the nuances of a client's modality functioning, the 35-item SPI was created.19 The reliability and validity of SPI were demonstrated in a study by Landes,26 who showed internal consistency and test-retest reliability, as well as concurrent validity, for the Affects, Sensations, Cognitions, and Interpersonal Relations modality scales through correlations with the MBTI. Recently, the SPI has been shown to have even higher reliability scores than previously demonstrated, and, through a correlation with the Vocational Preference Inventory, validity has been established for the Affects, Sensations, Imagery, and Interpersonal Relations modalities, with some indication of validity established for the Behaviors modality.27

Research has already demonstrated that therapists' theoretical stances are consistent with their own modality structures on the SPI22 and that clients' perceptions of their psychological difficulties can be similarly predicted from their SPI scores.24 The implications of these findings are that psychotherapists are likely to employ specific techniques that are consistent with their own modality structures and clients, likewise, are apt to see techniques consistent with their own modality structures as more pertinent to their issues. Herman23 has already demonstrated that when therapists and clients differ in their modality structures, early psychotherapy impact suffers. It is therefore intuitive to assume that psychotherapy outcome will suffer similarly. This is not a radical idea in psychotherapy research. McConnaughy28 discussed in detail the impact of a therapist's personality style on the style, form, and content of the psychotherapy practiced. Lazarus29 has recently reviewed the necessity of approaching the client on his or her own terms. Rogers had covered this same topic extensively for years previously,30 although employing a much more limited model.

This study was designed to explore the importance of therapist–client similarity from the holistic and comprehensive viewpoint of multimodal therapy. For this purpose, the SPI, the primary psychometric of multimodal therapy, was used to determine a measure of similarity. It was hypothesized that therapist–client similarity on the Multimodal Structural Profile Inventory would result in more successful psychotherapy outcomes than would therapist–client dissimilarity.

Methods

Subjects

Therapists were recruited by mailing letters to the directors of eight clinics in New York, New Jersey, Pennsylvania, and Indiana. Although initially 45 therapists agreed to participate, a total of only 19 therapist–client pairs did participate in this study. Therapists were primarily female (74%). The average age of the therapists was 32.8 years (SD = 6.8, range 23–46). Most of the therapists (74%) were still in advanced degree programs (Ph.D., Psy.D., Ed.D., or psychiatry residency), and they had an average of 3.96 years of experience (SD = 3.9, range 0–12). Half of the therapists were married (47%), and most were Caucasian (68%).

Clients were also primarily female (84%). Their average age was 27.0 years (SD = 8.6, range 18–49). Half of the clients were college students (47.4%). Several clients had master's degrees (26.3%), and a few had professional degrees. Of the nonstudents, some were employed and some were not. Most of the clients (74%) were single, 21% were married, and 5% were divorced. Most clients were Caucasian (74%).

Clients were excluded from participation if they were actively psychotic, actively substance dependent, or organically impaired. Clients had to be literate to the extent that they were capable of reading and answering the research questions in order to participate. Only clients 18 years or age or older were recruited for participation. Family and marital therapy clients were not recruited, so as to optimize concentration on the interactions between therapists and individual therapy clients. Each therapist was asked to participate only once with one individual client.

Procedures

Therapist–client similarity was determined through administration of the Multimodal Structural Profile Inventory, Version 3.19 Psychotherapeutic outcome was determined by use of the Brief Symptom Inventory31 (BSI). Correlations between the BSI and the Symptom Checklist–90 (SCL-90) are high enough that the BSI can be considered interchangeable with the SCL90, sharing the same reliability and validity characteristics.32 Concurrent validity for the BSI has been established through correlations between the SCL-90 and several other assessments, including the Minnesota Multiphasic Personality Inventory (MMPI),33 rendering the SCL-90 a frequently used outcome measure in psychotherapy research.9, 34,35

The Multimodal SPI19,22–24,26,28,36 is a 35-item self-report questionnaire that asks clients to rate their perceptions of their modality functioning according to a seven-point Likert scale. The SPI assays an individual's functioning in the realm of Behaviors (e.g., “I keep busy doing things”), Affects (e.g., “In making a decision, I often let my emotions be the key factor in determining what I should do”), Sensation, Imagery (e.g., “I am tuned in to my sensations: what I see, hear, taste, smell, and touch”), Cognitions (e.g., “I tend to plan things and think about them a great deal”), Interpersonal Relationships (e.g., “I have several close or intimate friends”), and Biological/Physical factors (e.g., “I follow good nutrition habits”). Scale scores can range from 5 (indicating poor functioning in the modality area or a preference against using that sphere of functioning) to 35 points (indicating high functioning in the modality area or a preference to use that sphere of functioning, that is, a dominant modality).

The BSI37 is a 53-item questionnaire that factors into nine clinical scales (Somatization, Obsessive-Compulsive, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism) and two summary scales, the General Severity Index (GSI) and the Positive Symptom Distress Index (PSDI). The GSI provides information about the general level of psychological symptomatology the client is experiencing, whereas the PSDI provides information about the severity of the specific symptoms the client endorses. All scales of the BSI can be analyzed by their raw scores (which range from 0, indicating an absence of psychopathological symptoms, to 4, indicating high levels of psychopathological symptoms) or can be converted to scaled scores, with a mean of 50 and a standard deviation of 10.

Both therapists and clients also completed a brief demographics questionnaire requesting information such as age, gender, race, occupation, and marital status. In addition, the therapists' demographics questionnaire requested philosophical orientation, degree sought, and years of psychotherapeutic experience.

Once therapists agreed to participate in this study, they were sent packets containing all the materials they would need. In addition to self-addressed return envelopes containing the intake materials and the session #12 questionnaires, therapists were provided with a detailed instruction sheet, two copies of their own consent form, the therapist SPI and therapist demographics questionnaire, two copies of the client consent form, and an “intake consent form.” The intake consent form, requesting only the therapist's name, the name of the clinic, and the therapist's signature to affirm that the client had signed his or her consent form, was included to ensure the client's anonymity. The experimenter thus did not receive the names of any of the clients. Therapists were instructed to complete and return their own consent form, SPI, and demographics sheet upon receipt of the research materials.

When therapists met with their next new client for the first time, they were instructed to explain the study and seek the client's consent for participation. Interested clients were then asked to read and sign both copies of the Client Consent Form, which described the study in detail. Although there were no enforcement mechanisms to ensure that therapists would participate with their next new random client, many of the participants were recruited through training clinics where clinicians saw only one or two clients per year. This was one of the major causes of the high dropout rate: many clients chose to not participate, regardless of the clinician's interest.

After the client signed the consent form, the therapist completed the Intake Consent Form (described above) and gave the client the Session #1 packet, which contained the BSI, the SPI, and the Client Demographics sheet in a self-addressed stamped envelope. Clients were instructed to complete and return these questionnaires immediately after the first session.

After the twelfth therapy session, the clinicians were instructed to remind their clients about the study and request that they complete the Session #12 packet, which contained the SPI and BSI in a self-addressed stamped envelope. Clients were instructed to complete these materials immediately after the twelfth therapy session.

According to the research protocol, therapists were instructed to administer exactly the same course of treatment that they would have followed had they not participated in the study. Clients were also informed that participation in this study would not alter the course or form of treatment that they would receive.

Statistical Analyses

Consistent with Cronbach and Glesser,38 therapist–client difference values were computed by using SPI modality scores corrected for elevation and scatter (D″2), converting them essentially to z-scores. The process begins by first obtaining a mean modality score [x = (B + A + S + I + C + I. + D.)/7], called the “profile elevation,” and a standard deviation of the modality scores, called the “profile scatter,” for both therapist and client. Each modality score is then subtracted from this mean score, correcting it for “elevation” (e.g., B′ = B – x). Modality scores (corrected for elevation) are then corrected for “scatter” by dividing them by the standard deviation of the modality scores (e.g., B″ = B′/SD). These steps are performed independently for both the therapist and client SPI scores. D″2 scores are computed by subtracting the client modality scores from the therapist modality scores, squaring the differences, and adding them together: D″2 = (B″t – B″c)2 + (A″t – A″c)2 + (S″t – S″c)2 + (I″t – I″c)2 + (C″t – C″c)2 + (I.″t – I.″c)2 + (D.″t – D.″c)2. The reader is referred to Cronbach and Glesser38 for a more in-depth description of the process and rationale for the use of this statistic. Using the D″2 statistic, the larger the obtained value, the greater the degree of dissimilarity (and thus the lesser the degree of similarity).

This conversion process prevented contamination of the dependent measure (the raw SPI scores). Uncorrected raw SPI modality scores have been found to be predictive of clients' symptomatology, as measured by BSI symptom scores, but SPI modality scores corrected for elevation and scatter have not.36 Thus, an additional strength of the SPI is that it can be used to roughly measure the nature and severity of a client's clinical symptoms.

In the analyses of therapeutic outcome, regression equations were computed by using the GSI scores from the session #12 BSI as the dependent measure of outcome. The independent measures in these equations were corrected therapist–client similarity scores (D″2) from the SPI and GSI scores from the Session #1 BSI (to obtain a pretreatment baseline). Only GSI scores were used in the analyses of outcome, rather than the nine different symptom clusters, in an attempt to avoid making a type I error. Because only 19 outcome packets were collected, it was determined that there were insufficient data for these nine separate analyses.

Results

Descriptively, therapist–client similarity scores (SPI D-scores corrected for elevation and scatter) were found to range from 1.11 to 4.89 (mean = 3.50, SD = 1.02), with larger numbers signifying greater differences between therapists and clients in modality functioning. Mathematically, the largest difference (D″2) possible between two individuals on the SPI is 5.29.

To demonstrate that this was a valid study of therapy and that the BSI was an appropriate instrument for examining outcome, t-tests for paired samples were computed on the GSI scores, comparing pretreatment to posttreatment. A significant decrease in symptomatology was observed (n = 18, t = 2.12, P = 0.04), from 1.18 (SD = 0.65, standard error = 0.15) before treatment to 0.93 (SD = 0.64, standard error = 0.15) after treatment. Therapist–client similarity was not found to be predictive of clients' initial levels of psychopathology in analyses of intake GSI scores (R = 0.22, R2 = 0.04, P = 0.09). This is consistent with findings that SPI scores corrected for elevation and scatter are not predictive of psychopathology.36

Outcome, as measured with the GSI, was not affected by the therapist's years of experience (R = 0.21, R2 = 0.04, P = 0.45) or by the therapist's status as a student or a professional (F = 0.69, df = 1,16, P = 0.41).

Analyses of the relationship between client–therapist similarity on the Multimodal SPI and psychotherapy outcome confirmed the experimental hypothesis. In the regression analysis, a significant relationship was observed between therapist–client similarity on the SPI and outcome as determine by pretreatment to posttreatment GSI scores (R = 0.79, R2 = 0.63, P = 0.03, β = 0.37). The more similar the therapist and client, the lower the degree of reported psychopathology at psychotherapy outcome.

Given the high dropout rate encountered in this study, some post hoc analyses were conducted to explore the relationship between therapist–client similarity on the SPI and dropout, as well as between several of the demographic variables and dropout. In these analyses, it was observed that dropout could not be predicted by D2, D′2, or D″2 statistics or by most therapist demographic data. When analyses of variance were computed exploring the relationships with the length of treatment (the number of sessions attended by the client), there was a surprising finding that the age of the therapist (F = 6.42, df = 1,45, P < 0.01) and the therapist's years of experience (F = 5.58, df = 1,45, P = 0.02) were predictive of early dropout from therapy. The surprising aspect was that it was the older, more experienced therapists who were more likely to have clients leave therapy prior to the twelfth session.

Discussion

When therapists and clients are more similar in their modality orientation, there appears to be a concomitant improvement in psychotherapy outcome associated with this degree of similarity. Why might this be the case? Does similarity lead to the enhancement of rapport? Does similarity improve and clarify communication? Does similarity make it easier for the therapist to choose interventions that will be more helpful to the client?

It has already been demonstrated that there is a relationship between a therapist's modality structure and his or her adherence to a particular camp of psychotherapeutic thought.22 And what is a treatment theory other than a template with which to make choices about technique selection?39 It would therefore be considered apparent that the therapist's modality structure plays an important role in determining the psychotherapeutic techniques likely to be used. However, this study did not follow therapy process, only outcome, and thus this conclusion cannot be drawn from the results of this study.

Psychotherapy is an active process. Regardless of the therapist's philosophical orientation, it is the application of administered techniques that results in change and symptomatic relief. These techniques must be communicated and perceived in order for them to have any true efficacy. Techniques can be administered (and perceived) by only two means: verbally and behaviorally. Similarity of modality structure between therapists and clients has already been demonstrated to be an important predictor of the establishment of rapport in early psychotherapy.23 Thus, even if therapists choose the same techniques and practice psychotherapy exactly the same regardless of their philosophical orientation, it appears that the techniques will be presented more clearly, be more “on target,” and have a greater degree of efficacy when therapists and clients are more similar in their modality orientation. Frank40 postulated that in order for psychotherapy to be effective, clients must perceive it as being effective. Because clients will be more likely to perceive and explain their own psychopathology in terms of their own modality functioning,24 this would suggest that when clients and therapists are similar in their modality functioning, clients will perceive the psychotherapeutic interventions used as being more “on target” and thus more effective and pertinent in treating their issues. Even if more dissimilar therapists are able to eventually modify their delivery, or even if more dissimilar clients are able to eventually translate or make use of what they receive from their therapist, the loss of productive therapy time apparently takes its toll in attenuated outcome levels. Regardless of the mechanism underlying the process, modality similarity between therapists and clients does appear to have a positive effect on the effectiveness of psychotherapy outcome.

It therefore appears that when therapists and clients have similar modality structures, not only will session impact be experienced by clients as more positive, arousing, engaging, and deep, but also clients will be likely to achieve a greater degree of symptomatic relief. There was some indication, based on the computed regression lines, that in the case of extreme dissimilarity, clients might even report increased symptoms after a course of psychotherapy. We would not expect this finding actually to be validated, since it is more likely that clients would probably terminate therapy (or be hospitalized) before worsening to the extent suggested by the regression line.

The results of this study suggest that despite the therapist's theoretical orientation, and regardless of the specific techniques employed in psychotherapy, the match between the therapist's and client's modality orientation will have significant implications for psychotherapy outcome. It was especially interesting to note that this phenomenon occurred regardless of the therapist's status as a student, or even his or her years of experience.

The conclusions of this study may seem to be somewhat attenuated and artificial because a measurement after 12 sessions was defined as “outcome.” After all, this was not truly “outcome”—clients were not terminated or considered “cured” at the end of 12 sessions. The decision to collect “outcome” data after the twelfth session was driven mostly by the fact that most of the clinics at which data were collected had a 12-session limit on treatment. The results of this study may not give a picture of total symptom reduction at the “close” of therapy, but they do give a clear idea of the speed and efficiency with which symptom reduction may occur as a function of the therapist–client modality match. Other weaknesses of this study concern the small sample size and the fact that most clinicians were still in training. It may be difficult to fathom generalizing from a subject pool of 19 therapist–client pairs to the field of psychotherapy in general. However, it must be noted that even with the small sample size, the observed effect was quite robust.

Some brief discussion of the high dropout rate is appropriate. In some cases, subjects dropped out of the study because the clients dropped out of therapy. In other cases, subjects dropped out of the study because the client or therapist “forgot” to participate. In other cases, subjects dropped out of the study because treatment goals were met earlier than session 12. It was expected, given the work of Mendelson and Geller,9 that therapist–client dissimilarity (D″2 scores) would predict dropout rates. That these dropout rates could not be so predicted was a surprising finding and is not easily explainable.

It may be beneficial in future research to examine this finding with a larger sample size, or perhaps with specific, homogeneous subject populations. Such more detailed analyses and explorations may provide even more understanding of the phenomenon of therapist–client matching.

Acknowledgments

The author deeply appreciates the contributions of the following to the preparation of this study: John A. Carpenter, Ph.D.; Cyril Franks, Ph.D.; Alan Wisenfeld, Ph.D.; and especially Arnold A. Lazarus, Ph.D., without whose guidance and assistance this project would not have been possible. This study was funded through the generosity of the American Psychiatric Association Division 29 President's Award, the Graduate School of Arts and Sciences of Rutgers—the State University of New Jersey, and the Indiana Association for the Advancement of Mental Health Research and Education.

References

1. Paul GL: Strategy of outcome research in psychotherapy. Journal of Consulting Psychology 1967; 31:109–118 [PubMed] [Google Scholar]

2. Norcross JC: All in the family? On therapeutic commonalities. Am Psychol 1981; 36:1544–1545 [Google Scholar]

3. Lambert MJ: Implications of psychotherapy outcome research for eclectic psychotherapy, in Handbook of Eclectic Psychotherapy, edited by Norcross JC. New York, Brunner/Mazel, 1986, pp 436–462

4. Smith ML, Glass GV, Miller TI: The Benefits of Psychotherapy. Baltimore, MD, Johns Hopkins University Press, 1980

5. Hunt DD, Carr JE, Dagadkis CS, et al: Cognitive match as a predictor of psychotherapy outcome. Psychotherapy 1985; 22:718–721 [Google Scholar]

6. Foon AE: Locus of control and clients' expectations of psychotherapeutic outcome. Br J Clin Psychol 1986; 25:161–171 [PubMed] [Google Scholar]

7. Dormaar M, Dijkman CIM, de Vries M: Consensus in patient–therapist interactions. Psychother Psychosom 1989; 51:69–76 [PubMed] [Google Scholar]

8. Cummings AL, Martin J, Hallberg E, et al: Clients' and counselors' experiences of important events in counseling. Paper presented at the 99th annual convention of the American Psychological Association, San Francisco, CA, 1991

9. Mendelson GA, Geller MH: Similarity, missed sessions, and early termination. Journal of Counseling Psychology 1967; 14:210–215 [Google Scholar]

10. Baron J, Baron JH, Barber JP, et al: Rational thinking as a goal of therapy. Journal of Cognitive Psychotherapy: An International Quarterly 1990; 4:293–302 [Google Scholar]

11. Persons JB: The process of change in cognitive therapy: schema change or acquisition of compensatory skills? Cognitive Therapy and Research 1993; 2:123–137 [Google Scholar]

12. Mahoney MJ, Norcross JC: Relationship styles and therapeutic choices: a commentary on the preceding four articles. Psychotherapy 1993; 30:423–426 [Google Scholar]

13. Kemp D: Routenizing art: implications of research with the A-B scale for the practice of psychotherapy. Journal of the American College Health Association 1970; 18:238–240 [PubMed] [Google Scholar]

14. Seidman E: A and B subject–therapists' responses to videotaped schizoid and neurotic prototypes. J Consult Clin Psychol 1971; 37:201–208 [PubMed] [Google Scholar]

15. Beutler LE, Johnson DD, Neville C, et al: Accurate empathy and the A-B dichotomy. J Consult Clin Psychol 1972; 38:372–375 [PubMed] [Google Scholar]

16. Green M: The effectiveness of A and B college males as interviewers, as response-biasers, and as verbal conditioners with schizophrenic and neurotic patients. Doctoral dissertation, University of Kentucky, Lexington, KY. Dissertation Abstracts International 1971; 33-02(B):0915

17. Razin A: The A-B variable: still promising after 20 years? In Effective Psychotherapy: A Handbook of Research, edited by Gurman A, Razin A. New York, Pergamon, 1978, pp 291–324

18. Beutler LE, Clarkin JF: Systematic Treatment Selection. New York, Brunner/Mazel, 1990

19. Lazarus AA: The Practice of Multimodal Therapy (updated edition). Baltimore, MD, Johns Hopkins University Press, 1989

20. DeRubeis RJ, Feeley M: Determinants of change in cognitive therapy for depression. Cognitive Therapy and Research 1990; 14:469–482 [Google Scholar]

21. Crits-Cristoph P, Cooper A, Luborsky L: The accuracy of therapists' interpretations and the outcome of dynamic psychotherapy. J Consult Clin Psychol 1988; 56:490–495 [PubMed] [Google Scholar]

22. Herman SM: Predicting psychotherapists' treatment theories by multimodal structural profile inventory responses: an exploratory study. Psychotherapy in Private Practice 1992; 11(2):85–100

23. Herman SM: Therapist–client similarity on the Multimodal Structural Profile Inventory as a predictor of early session impact. J Psychother Pract Res 1997; 6:139–144 [PMC free article] [PubMed] [Google Scholar]

24. Herman SM, Cave SK, Kooreman HE, et al: Predicting clients' perceptions of their symptomatology by Multimodal Structural Profile Inventory responses. Psychotherapy in Private Practice 1995; 14(1):23–33

25. Lazarus AA: The Practice of Multimodal Therapy. New York, Macmillan, 1981

26. Landes AA: Development of the Structural Profile Inventory. Psychotherapy in Private Practice 1991; 9(2):123–141

27. Herman SM: A demonstration of the validity of the Multimodal Structural Profile Inventory through a correlation with the vocational preference inventory. Psychotherapy in Private Practice 1992; 11(3):71–80

28. McConnaughy EA: The person of the therapist in psychotherapeutic practice. Psychotherapy 1987; 24:303–314 [Google Scholar]

29. Lazarus AA: Tailoring the therapeutic relationship, or being an authentic chameleon. Psychotherapy 1993; 30:404–407 [Google Scholar]

30. Rogers CR: Client-centered therapy, in American Handbook of Psychiatry, vol 3, edited by Arieti S. New York, NY, Basic Books, 1959, pp 183–200

31. Derogatis LR: Brief Symptom Inventory. Baltimore, MD, Clinical Psychometric Research, 1975

32. Derogatis LR, Melisaratos N: The Brief Symptom Inventory: an introductory report. Psychol Med 1983; 13:595–605 [PubMed] [Google Scholar]

33. Derogatis LR, Rickels K, Rock AF: The SCL-90 and the MMPI: a step in the validation of a new self-report scale. Br J Psychiatry 1976; 128:280–289 [PubMed] [Google Scholar]

34. Steer RA, Platt JJ, Rainert WF, et al: Relationships of SCL-90 profiles to methadone patients' psychosocial characteristics and treatment response. Multivariate Experimental Clinical Research 1989; 9:45–54 [Google Scholar]

35. Talley PF, Strupp HH, Morey LC: Matchmaking in psychotherapy: patient–therapist dimensions and their impact on outcome. Journal of Counseling and Clinical Psychology 1990; 58:182–188 [PubMed] [Google Scholar]

36. Herman SM: The diagnostic utility of the Multimodal Structural Profile Inventory. Psychotherapy in Private Practice 1994; 13(3):55–62

37. Derogatis LR, Spencer PM: The Brief Symptom Inventory (BSI): Administration, Scoring and Procedures Manual, I. Baltimore, MD, Clinical Psychometric Research, 1982

38. Cronbach LJ, Glesser GC: Assessing similarity between profiles. Psychol Bull 1953; 50:456–473 [PubMed] [Google Scholar]

39. Messer S: Behavioral and psychoanalytic perspectives at therapeutic choice points. Am Psychol 1986; 41:1261–1272 [PubMed] [Google Scholar]

40. Frank JD: Persuasion and Healing: A Comparative Study of Psychotherapy (revised edition). Baltimore, MD, Johns Hopkins University Press, 1973

Articles from The Journal of Psychotherapy Practice and Research are provided here courtesy of American Psychiatric Publishing