The Therapist as a Neglected Variable in Psychotherapy

Research
Sol L. Garfield, Washington University

Research on psychotherapy has increased in both quan- forms of psychotherapy have been developed (Barlow &
tity and quality over the past 30 years and has tended Cerny, 1988; Beck, Rush, Shaw, & Emery, 1979; Kler-
to focus on the evaluation of outcome. The major em- man, Weissman, Rounsavdle, & Chevron, 1984; Lubor-
phasis has been on studies comparing different forms sky, 1984; Strupp & Binder, 1984). Thus, therapists can
of psychotherapy. The recent emphasis on training
be trained to deliver the specific therapy as described in
the manual and can be monitored so that they do not
manuals has reinforced this pattern. On the other hand,
“drift” fiom the prescribed procedures. This was the pro-
the importance of the therapist’s contribution to out-
cedure followed in the large-scale National Institute of
come and the related matter of therapist variability
Mental Health Treatment of Depression Collaborative
have been given inadequate attention. These issues are
Research Program, where the psychotherapists were
discussed and evaluated in the articles that follow. trained and supervised by experts of the two psychothera-
Key wods: psychotherapy, effectiveness, therapist pies evaluated (Ellun, Parloff, Hadley, & Autry, 1985).
contribution. [din Psycho1 Sci Prac 4:4043, 19971 Such attempts to clearly specifjl the theoretical and
technical aspects of the forms of psychotherapy can be
The amount of research conducted on psychotherapy has viewed, perhaps, as the epitome of the emphasis on the
increased noticeably over the past 30 years. In addition to importance of the form or type of psychotherapy. There
an increase in the quantity of the research there has also is little question that the major focus in psychotherapy
occurred an increase in the sophistication of research research has been on comparative outcome studies in
designs and methods of analysis. Uncontrolled and sub- which specified forms of psychotherapy have been com-
jectively appraised studies of psychotherapy outcome are pared with a control group or with one or more other
currently few and far between. Most studies tend to have forms of psychotherapy, for example, psychodynamic,
a reasonably appropriate control group and to use a t least cognitive, behavioral, client-centered, and the like. The
some standardized measures of outcome. Furthermore, whole issue about the effectiveness of psychotherapy has
there has been evident also a movement to evaluate revolved around such studies.
groups of clinical cases meeting certain criteria for spe- Without question, such comparative studies have
cific disorders or types of psychopathology instead of a played an important role in providmg research data to
group of patients representing a variety of disorders. support the effectiveness of psychotherapy and to answer
Other improvements and developments have been the attacks of its critics. This is a significant contribution
apparent in more recent years. As a means of increasing that should not be minimized. However, such research
the internal validity of the comparative studies of forms has tended to overemphasize the form or type of psycho-
of psychotherapy, manuals for conducting the specified therapy and has for the most part slighted other poten-
tially important aspects of psychotherapy. One such
Address correspondence to Sol L. Garfield, Ph.D., Department matter concerns the importance of the contribution of
of Psychology, Wastungton University, Campus Box 1125, #1 the individual psychotherapist to the process and out-
Brookings Dr., St. Louis, MO 63130-4899. come of psychotherapy. The variability in the perfor-

0 1 9 9 7 AMERICAN PSYCHOLOGICAL ASSOCIATION D l 2 * 0969-5893/97/55.00 40

mance and skill of the individual psychotherapists approach to research on psychotherapy outcome. The
involved in the comparative stuhes has for the most part more recent development of specific therapy manuals for
been minimized or overlooked. The most recent example specific psychiatric disorders has actually increased this
is the recent report of the American Psychological Asso- emphasis. The matter of individual differences among
ciation (Division of Clinical Psychology) Task Force on psychotherapists of the same orientation and its potential
Promotion and Dissemination of Psychological Proce- importance for outcome, however, has remained a rather
dures (1995). Again, the emphasis in this report is placed neglected topic and area of research.
on the type of therapy, and relatively little attention paid There are also some other plausible explanations for
to the potential variability among psychotherapists. the lack of research on differences in effectiveness among
Although most psychotherapists recognize that not all psychotherapists. Studies of this type involve individuals
therapists are equally effective, relatively few studies have and conceivably could influence their h t u r e careers.
devoted any significant attention to this important aspect Such matters obviously cannot be taken lightly. Results
of psychotherapy. Instead, as pointed out years ago by of such potential studies also may reflect negatively on a
Kiesler (1966, 1971), we have appeared to support the particular institution or clinical setting. In one instance,
uniformity hypothesis-all therapists of a given orienta- for example, a colleague of mine was not given permis-
tion perform equally well. This is a very interesting phe- sion to publish a study he had made of therapists’ perfor-
nomenon since in every facet of human activity, we are mance because the results of one therapist were
quite aware ofwide differences in performance, skill, per- exceedingly poor. Although such studies can be very
sonality, and knowledge. Anyone who has ever taught a threatening to individual therapists, and one can empa-
college course has encountered wide differences in the t h z e with them, one can also raise concerns about the
performance and interest of the students, and as psycho- welfare of patients treated by less than effective therapists.
therapists, we have also encountered wide differences in At least some communities keep records of excessive
patients in every aspect of our work, and no examples fatalities resulting fiom surgery and on some occasions
need be offered here. may limit the surgeon’s right to practice. Although most
Why, then, have there been so many studies on type likely we would not view psychotherapy as exactly the
of psychotherapy and so few on therapist variability or equivalent of major surgery, we should not just casually
the contribution of the individual therapist? There are at dismiss t h s issue.
least a few possible answers to this query. It would appear It is also interesting to note that at various times in the
that even during the early development of psychoanalysis past a number of desirable personal requisites have been
the trend toward emphasizing different theoretical specified for the practice of psychotherapy or psychoanal-
emphases was not uncommon. Thus, relatively early, ysis. At the Boulder conference on the training of clinical
Jung and Adler went their separate ways and developed psychologists, for example, a list of 15 recommended
their own approaches-to be followed later by Rank, characteristics were considered including superior ability,
Homey, and others. The emphasis clearly was on theoret- interest in persons as individuals, insight into one’s own
ical differences and type of therapy. personality characteristics, sensitivity to the complexities
This pattern of creating new therapeutic approaches of motivation, tolerance, and the ability to establish warm
and of therapists identifying themselves with a particular and effective relationships with others (Raimy, 1950). A
approach has continued unabated. The different forms of few years later, Holt and Luborsky (1958) published an
psychotherapy now number in the hundreds (Herink, even more spectacular list of 25 desired quahies for psy-
1980; Karasu, 1986), and creative efforts in this area have choanalysts. Thus, although very high personal qualifica-
not yet ceased. For whatever reasons, many people have tions have been recommended for the practice of
tended to get emotionally involved and identified with a psychotherapy, actual appraisals of the effectiveness of the
particular orientation. Consequently, with the emphasis therapists trained have been few and fir between. How-
on type of psychotherapy, and with this type of emphasis ever, when we do sometimes devote attention to therapist
also present in many introductory textbooks on psycho- variability, we can note important differences among
therapy, it is not surprising that evaluations of specific those appraised, and negative outcomes in therapy have
types of psychotherapy became the predominant also been secured (Lambert & Bergin, 1994).

THERAPIST-A NEGLECTED VARIABLE GARFIELD 41

 Some years ago, I conducted a study of continuation Clearly, it is timefor us to discard the ungormity myths in the
in psychotherapy that focused on the first therapy inter- field ofpsychotherapy, and tofocus much more on the qualities
view as a potential predictor (Garfield, Affleck, & Muffly, and behaviors OJ the participants as they are manijkted in the
1963). My colleagues and I taped the initial therapy inter- interactions we callpsychotherapy. Only in this way can we secure
views that each of six therapists had with four new more meaningful data on those processes which may lead to really
patients. Our main focus of interest was on the types of positive outcomes in psychotherapy. (Garfield, 1981, p p .
interactions that occurred during the interviews and their 3 05-306)
relationship to continuation. At the completion of the
study and prior to any knowledge of the results regarding In the present series of articles, therefore, the focus is
attrition, the three judges in the study ranked the six ther- on one of these players, the psychotherapist. Each of the
apists on their overall effectiveness as therapists. These authors in the following four articles will discuss findings
rankings were highly reliable. M e r the data were secured and issues pertaining to the psychotherapist as a neglected
on patient continuation in therapy, we compared the variable in psychotherapy. These presentations wdl be
results for each of the six therapists. The findings were of followed by a brief summation and commentary prepared
real interest: “The two most favorably rated therapists by Allen Bergin.
each kept three out of four of their patients, whereas the
two least fivorably rated therapists each kept only one out A C K N O W L E D G M EN T
of four assigned patients” (Garfield et al., 1963, p. 477). This work was originally presented in S. L. Garfield (Chair),
The two remaining therapists each had two remainers The Psychothwupist as a Neglected Variable in Psychotherapy Research,
and two terminators. Clearly these therapists differed a symposium conducted at the annual meeting of the Society
noticeably on this criterion, but nothing more was made for Psychotherapy Research, Vancouver, Canada,June, 1995.
of this and the confidentiahty of the participants was
definitely maintained. REFERENCES
In a more recent analysis of outcomes in four studes, American Psychological Association Task Force on Promotion
Luborsky et al. (1986) compared the amount of variance and Dissemination of Psychological Procedures. (1995).
that could be attributed to the therapeutic t e c h q u e s to Training in and hemination of empirically-validated psy-
the amount that could be attributed reasonably to the chological treatments: Report and recommendations. The
Clinical Psychologist, 48, 3-23.
therapist. They concluded that “the size of therapists’
Barlow, D. H.,& Cemy, J. A. (1988).Psychological treatment of
effects generally overshadowed any dfferences between
panic. New York Gullford Press.
different forms of treatment in these investigations”
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979).Cog-
(Luborsky et al., 1986, p. 509). Additional findings ofthis nitive thwupy of depression. A treatment manual. New Yo&
type are referred to in the articles that follow. Guilford Press.
As indicated previously, there has been a noticeable FJkin, I., ParloK M. B., Hadley, S. W., & Autry, J. H. (1985).
increase in the quantity and quahty of research on evalu- NIMH Treatment of Depression Collaborative Research
ating outcome in psychotherapy. Most of t)lls research has Program. Archives ofGenerul Psychiutty, 42, 305-316.
focused on comparing different forms of psychotherapy, Garfield, S. L. (1981).Evaluating the psychotherapies. Behuvior
and the results have generally been positive. There would Thprapy, 12, 295-307.
appear to be a greatly dimmshed need for such research Garfield, S.L. (1990)Issues and methods in psychotherapy pro-
in the hture. O n the other hand, our knowledge of the cess research. Journal oj Consulting and Clinical Psychology,
variables that make for positive change as a result ofpsy- 58, 273-280.
Garfield, S. L., Affleck, D.C., & Muffly, R. (1963).A study of
chotherapy is s d quite limited (Garfield, 1990). As a
psychotherapy interaction and continuation in psychother-
result, there is at present a greater awareness of the need
apy.Journal ofClinicul Psychology, 19, 473-478.
for high-quahty process research to enlarge our under- Herink, R.(Ed.). (1980).The psychotherapy handbook: The A to
standmg of the actual therapeutic process. It is readily Zguide to moz thun 250 diferent therapies in use today. New
apparent that the key players in this process, the client York Meridan.
and the therapist, are of some importance in how this Holt, R.R., & Lubonky, L. (1958).Pwsonalitypattems ofpsyhi-
process develops and in the outcomes secured. atrists (Vol. 1). New York Basic Books.

CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V4 N l , SPRING 1997 41

Karasu, T. B. (1986). The specificity versus nonspeclficity Luborsky, L. (1984) Principles of psychoanalytic psychotherapy: A
dilemma: Toward identifjmg therapeutic change agents. manual for supportive-expressive treatment. New York: Basic
Americanjournal ofPsychiatry, 143, 687-695. Books.
Kiesler, D. J. (1966). Some myths of psychotherapy research Lubonky, L., Crits-Christoph, P., McLellan, A. T., Woody, G.,
and the search for a paradigm. Psychological Bulletin, 65, Piper, W., Liberman, B., Imber, S., & Pilkonis, P. (1986).
110-136. Do therapists vary much in their success? Findings &om four
Kiesler, D. J. (1971). Experimental designs in psychotherapy. In outcome studies. American Journal of Orthopsychiatry, 56,
A. E. Bergin & S. L. Garfield (Eds.), Handbook ofpsychother- 501-512. .
apy and behavior change @p. 36-74). New York: Wiley. Raimy, V. (Ed.). (1950). Training in clinicalpsychology. New York
Klerman, G. L., Weissman, M. M., Rounsaville, B. J., & Chev- Prentice-Hall, 1950.
ron, E. S. (1984). Interpersonal therapy of depression ( I . P T ) . Strupp, H. H. & Binder, J. L. (1984). Psychotherapy in a new key:
New York: Basic Books. A guide to time-limited dynamic psychotherapy. New York:
Lambert, M. J., & Bergin, A. E. (1994). The effectiveness of Basic Books.
psychotherapy. In A. E. Bergin & S. L. Garfield (Eds.),
Handbook of psychotherapy and behavior change (4th ed., pp.
143-189). New York: Wiley. Received October 24, 1995; accepted June 17, 1996.

THERAPIST-A NEGLECTED VARIABLE * GARFIELD 43