average effect size increased from .85 to .93 of a stan-
dard deviation unit. This figure would be analogous
to reducing an illness or death rate from 66% to 34%!
Table 11-4 presents a summary of Smith et al.’sfind-
ings for several major forms of psychotherapy.
Process Research
So far, we have focused on the outcomes of thera-
peutic intervention, along with many of the patient
and therapist variables that may affect those out-
comes. Other investigators, however, have addressed
the specific events that occurduringtherapy in the
course of the interaction between therapist and
patient. This is calledprocess research. The Rogerians
were pioneers in this area and commonly conducted
studies relating, for example, the amount of therapist
talk in a given session to client spontaneity or the
effects of therapist clarifications and restatements on
the client’s taking responsibility for the progress in a
given session.
For a long time, therapy investigators were split
into two camps (Beutler, 1990): those who did pro-
cess research and those who did outcome research.
Those who focused on outcomes criticized process
research as failing to show that processes internal to
therapy were correlated with outcome and, there-
fore, as not worthy of serious consideration. Indeed,
it sometimes seemed as if the process was misleading
in predicting outcomes.
But process-oriented researchers felt there
should be relationships between outcome and the
processes that occur during therapy (Strupp, 1971).
One way of doing this kind of research is to film or
tape actual therapy sessions (Hill & Lambert, 2004).
This kind of research has increasingly begun to
show substantive relationships between what hap-
pens during therapy and ultimate outcome
(Orlinski, Ronnestad, & Willutzki, 2004). For
example, across studies, one of the strongest associa-
tions found is between therapeutic bond (the
degree to which client and therapist are invested
in, communicate with, and affirm each other) and
outcome (Orlinski et al., 2004). This highlights the
importance of a strong, positive working relation-
ship in psychotherapy.
Other consistent findings concerning the rela-
tionship between measures of therapeutic process
and outcome include (Orlinsky et al., 2004):
■ Therapist competence and adherence to treat-
ment protocol are associated with positive
outcome;T A B L E 11-4 Average Effect Size (ES) and Percentile Equivalent for Select Forms
of Psychological Intervention
Type of Therapy ES Percentile Equivalent
Psychodynamic 0.69 75%
Client-centered 0.62 73%
Gestalt 0.64 74%
Rational-emotive therapy (RET) 0.68 75%
Non-RET cognitive therapies 2.38 99%
Systematic densensitization 1.05 85%
Behavior modification 0.73 77%
Cognitive-behavioral therapy 1.13 87%
Undifferentiated counseling 0.28 61%
All forms of psychological intervention 0.85 80%
NOTE:Percentile equivalentindicates the percentage of those not receiving treatment whose outcome is exceeded by those receiving the treatment in question.
SOURCE: Adapted with permission fromThe Benefits of Psychotherapy, by M. L. Smith, G. V. Glass, and T. I. Miller. Copyright © 1980 by Johns Hopkins
University Press.
PSYCHOLOGICAL INTERVENTIONS 337