THERAPIST VARIABLES

Issues of gender bias in therapy (Jones & Zoppel, 1982).

In study 1 Jones & Zoppel employed a 2 (gender of T, M vs. F) X 2 (gender of P, F vs M) design evaluating outcome ratings. The study employed 140 T and 160 P. On pre-post ratings of change made by M T and F T, M P and F P were rated equally improved.

F T rated all clients as more improved than M T did on: Happiness, Ability to Enjoy Life, Ability to get along with Spouse & Kid, and Ability to Handle Emotional Problems.

F T rated T P as sig. More improved than M T rated F P on: Symptoms, Ability to Enjoy Life, Relationship with spouse & Kid, and relationship with other relatives.

M T rated F P as sig. Less improved on Happiness than M T rated M P.

On process measures in Study 1, M T were more critical than F T in describing all P, with a notable tendency for the criticism to be harsher with F P.

Ratings of M P

Adjectives used more by F T than M T

intelligent, resentful, mild, immature, steady, resourceful, praising, distrustful, cold, appreciative

Adjectives used more by M T than F T

uninhibited sympathetic, peaceable, sentimental, cowardly, commonplace, preoccupied

Ratings of F P

Adjectives used more by F T than M T

capable, emotional, healthy shy, understanding, efficient, honest, confused, strong, intelligent

Adjectives used more by M T than F T

wary, simple, affected, commonplace, superstitious, conventional, awkward, cautious, conceited, temperamental

Study 2 used data from the Patients’ perspective. There was an overlap of some clients and therapists in Studies 1 and 2.

Outcome ratings by P: There were 2 main effects for gender of T. All patients of F T had: a) More energy to do things, and b) said that therapy fit their expectations (than P of M T).

There were 3 main effects for gender of P: F P rated themselves as more improved on a) Happiness, b) ability to enjoy life, and c) ability to handle personal problems.

THERE WERE NO OUTCOME INTERACTIONS IN STUDY 2, suggesting that there was no effect of gender bias by T on P’s perception of outcome.

Process rating by P: 1 main effect for gender of T. All patients viewed F T as more effective in the therapeutic alliance than M T.

One main effect for gender of P: F P viewed therapy as more of a Neg. Experience than M P. F P felt more depreciated.

One interaction of T X P: Same gender pairings were more likely to view their T as neutral, detached, or non-directive than opposite gender matches.

STRUPP AND ANDERSON (1979.)--The limitations of therapy manuals

Advantages of manual

better specify the independent variable

decrease therapist variance

adherence may enhance outcome

a useful method for outlining general principles

Limitations

hard to implement with “experienced therapists”

does not reflect “real life therapy”

restricting the therapist may be detrimental to outcome.

“The therapist’s demographic characteristics, attitudes, values, professional affiliations and techniques have little to do with the ability, as a person, to relate comfortably with others, to serve as a model for adult living, to be sensitive to subtle... dynamics, while appropriately maintaining professional distance. The techniques used by the therapist who successfully embodies these qualities are not easily captured in manualized form.”

Rigid reliance on technical rules may interfere with full attention to the patient’s communications.

“The process of acquiring new therapeutic skills may involve an extended period of disorganization and awkwardness.”

HANDBOOK CHAPTER ON THERAPIST VARIABLES (1994), by Beutler, Machado, & Neufeldt. New information in the (2004) chapter by Beutler et al. is presented in bold. The new information comes from studies published between 1990 and 2000.

General conclusions:

A). Benefit is more closely related to identity of the therapist than the type of therapy.

B). Some therapists in all approaches are consistently better.

C). Some therapists produce consistently neg. Results.

The review classified therapist variables into 4 categories:

1. Objective, cross-situational: now called Observable Traits:

Age of T: A weak relationship to outcome, if anything, similarity helps. No relationship

Gender: If anything, Female Therapists may be more helpful. Same gender matchups facilitate process, especially if T presents a non-stereotyped gender-role viewpoint. Less consistent effects on outcome. One meta-analysis shows d = .04, p < .05 in favor of female therapists. Beutler et al. find r = .01

Ethnicity: ethnic matches promote continuation, little or no effect on outcome for continuers. Five improvement studies and 3 dropout studies find sig. weak effects in support of ethnic match and improvement, r = .02, p < .05

2. Subjective, cross-situational: Now called Inferred Traits:

Personality of T: No replicated main effects of single personality variables on outcome.

Emotional well-being of T: Pos. Related to Outcome.

Personal therapy: No relationship to outcome.

Attitudes, values, and beliefs: For both religious and general values, there are better outcomes when T’s values are consistent or converging with P, or when T is able to accept P’s values. The same is true for cultural attitudes and gender-role orientation.

Little recent research has been done in this area.

3. Subjective, therapy-specific variables: Now called Inferred States:

Rogerian variables: Ps’ pos. Perceptions of T’s facilitative attitudes enhance outcome moderately. Ps’ ratings of T predict outcome better than independent observer’s ratings of T, suggesting that P may contribute to this helping alliance.

Social Influence Attributes: Ts’ attractiveness, trustworthiness, and expertise facilitate outcome.

Therapist expectations: congruence of expectation between T and P (ex., role induction interviews) enhance outcome.

The quality of the therapeutic relationship is consistently related to outcome.

4. Objective, therapy-specific variables: Now called Observable States:

Therapeutic orientation: No firm conclusions, no strong evidence for superiority of any orientation.

Professional discipline: NIMH Depression Study shows greater proportion of Ph.D.s (60%) in more effective therapist group, greater proportion of M.D.s (83%) in the less effective therapist group. The three most effective therapists were M.D.s, however, (Blatt et al., 1996)

Professional background: No clear picture that level of training, amount of experience, or discipline matters ONE POSSIBLE EXCEPTION: A meta-analysis of Weisz, Weiss, & Alicke (1987) suggested that Prof > Paraprofessionals with over controlled kids. Beutler et al. report a correlation of r = .48 between experience level and outcome in the Blatt et al. (1996) analysis of the NIMH Depression data. In this same report, Blatt et al. describe the no. of years experience (SD) in the three groups: More effective, 12.0 yrs. (7.5); Moderate effectiveness, 10.9 yrs. (8.2); and Less effective, 11.2 yrs (5.8) as not being significantly related to outcome.

“An emerging body of evidence suggests that the use of professional therapists may be contraindicated under certain circumstances…among those who have been subjected to traumatic stress.” (p. 238). Overall these findings tend to cast doubt on the validity of the suggestion that specific training in psychotherapy, even when un-confounded with general experience, may be related to therapeutic success or skill.” (p. 239)

Therapeutic styles: Using the 2 dimensions of Interpersonal Styles in Leary’s (orbiting the earth) Circumplex Model (i.e., Friendly-Unfriendly, and Dominance-Submission); similar levels of friendliness and contrasting levels of dominance were associated with better outcomes than other matches. Friendly therapist behaviors have been consistently related to better outcomes.

Verbal style: Topic changing promotes premature termination

Non-verbal style: no outcome data

Use of therapy manuals: Decreases variability across therapists. Little evidence that manuals increase outcomes in clinically relevant samples.

Therapist skill: Moderately related to outcome

Directiveness: Seems to be negatively related to outcome in dynamic therapy. With depressed clients who are resistance prone (defensive) directive < non-directive, with low resistance (non-defensive) clients directive > non-directive. This ATI continues to receive research support.

Insight-oriented versus symptom focused treatments: Beutler et al. conclude that for clients that are self-reflective, introverted, and introspective; insight-oriented treatments work better. For clients that are impulsive, aggressive, and undercontrolled; symptom-focused treatments work better.

Self-disclosure: No relationship. A small significant relationship

Interpretations: Questionable relationship with outcome. Ditto