until after the study is completed; in anattention only
control group, patients meet regularly with a clinician,
but no“active”treatment is administered. As much as
possible, patients in the treatment and control groups
are matched on variables that might be related to out-
come, such as gender, age, diagnostic status, and sever-
ity of symptoms before the study began, prior to being
randomly assigned to groups. Assessments ofpatient
functioning(symptoms of psychopathology) are con-
ducted in parallel fashion for both treatment and con-
trol participants. At the very least, assessments are
obtained at the beginning of the study, at treatment
completion, and possibly at some period of time after
treatment is terminated (follow-up). This design
allows a comparison of the two groups at treatment
completion and follow-up as well as an evaluation of
the amount of change (if any) within each group.
Following are some of the research considera-
tions that help shape the meaning and generality of
research findings on therapeutic outcomes.
- What is the sample? Are the patients voluntary,
or were they subtly or overtly coerced into
therapy (e.g., prisoners vs. private practice
patients). Were the therapists experienced, or
were they neophytes? Were they psychoana-
lysts, or were they behaviorists? Were the
patients“real”patients, or were they recruited
by a newspaper ad requesting paid volunteers
for an analog study on the“treatment of snake
phobias”? Undoubtedly, the answer to each of
these questions (and others like them) will
determine how researchers can interpret their
results. There are no absolute findings, only
findings relative to the sample and to the con-
ditions of the given study. - What relevant patient variables were con-
trolled? Unfortunately, one cannot hope to
provide a control group that is exactly the same
as the treatment group. This being the case,
how close did the study come to controlling
relevant factors? If the waiting list or attention
only control group was not identical with the
patient group, in what ways did it differ? Were
the presenting complaints of the patient group
all basically alike, or was there diversity? Was
the control group similar to the treatment
group with respect to demographic factors,
personality, knowledge about therapy, and
expectations for help?- What were theoutcome measures? Were the
outcome measures identical for every patient
and control, or were they“tailored”to meet
the idiosyncratic situation (goals, hopes, and
expectations) of each patient? Was a single
outcome measure used, or were multiple
measures employed? Were the measures non-
reactive or unobtrusive, or were they measures
that, by their very character, might reflect
things other than what they were supposed to
be measuring? Unfortunately, there is currently
no consensus about which outcome measures
should be used in psychotherapy research
studies. This makes comparisons across studies
more difficult. - What was the general nature of the study? The
effects of therapy can be evaluated in a variety
of ways. So far, we have concentrated on
experimental studies. Other methods include
case studies, clinical surveys, correlational
studies, and analog studies. Each type has
characteristic strengths and weaknesses. For
example, a case study can offer a richness of
detail and a fountain of hypotheses that may be
far more valuable than an experimental study
or an analog study. But a case study has anN
(i.e., sample size) of 1, and how far can one
generalize from one patient? Analog studies
offer great potential for controlling relevant
variables, avoiding ethical problems with no-
treatment groups, and collecting a satisfactory
number of participants. But how close to real-
ity is an analog study? As we have had occasion
to remark before, each method offers unique
advantages and characteristic limitations. How
researchers choose to proceed must be deter-
mined by what they seek to learn and what
they can tolerate in the way of limitations.
Perhaps the best hope is that numerous good
investigators will decide to follow diverse
research paths.
PSYCHOLOGICAL INTERVENTIONS 335