certain patient characteristics for purposes of predic-
tion. However, it may not be possible to uncover
those characteristics without extensive interviewing
or some in-depth interpretation based on test
results. As a further example, several facets of a
patient’s life history data may suggest to a sensitive
clinician that the patient could be prone to making
violent sexual assaults on women. Although clinical
psychology does not have a reliable regression
equation to predict such assaults, the data uncov-
ered by an astute clinician could be important.
Thus, at present, certain data may be discoverable
only through extensive clinical investigation. Pre-
dictive formulas work best when test data are avail-
able. Sometimes, however, tests of the right sort
simply do not exist. When dealing with rare events
(e.g., suicide), the frequency of occurrence is so low
that clinicians cannot develop adequate equations
for them. But rare or not, such events are impor-
tant, and they must be dealt with by clinical
judgment.
Finally, many people would argue that the
power to predict specific outcomes is not the only
goal of science; rather, understanding and describ-
ing phenomena are the overriding goals. Although
there may be some validity to this argument, all too
frequently it can become a rationalization for using
vague terminology and applying equally vague cri-
teria, as noted earlier. The counterargument would
assert that when description and understanding are
couched in explicit terms, with clear-cut referents
and criteria, then prediction will be a natural by-
product.
Comparing Clinical and Actuarial Approaches
Over the years, many studies have compared the
relative accuracy of clinical and actuarial methods
(Garb, 2005; Grove, Zald, Lebow, Snitz, & Nelson,
2000). Let us now examine some of that work.
Comparison Studies. Sarbin (1943) contrasted
the prediction of academic success for college fresh-
men made by a clerk employing a regression equation
with the predictions made by several counselors. The
regression equation predictors were aptitude test
scores and high school rank. The counselors had
available the two preceding sources of data (but with-
out their mathematical weighting), vocational interest
scores, interview data, and biographical data. Sarbin
(1943) found that the counselors were no better
than the regression equation in their predictions,
even though they had the benefit of much more
information.
Meehl (1954) surveyed a number of the studies
available on clinical versus statistical prediction and
concluded that in“all but one...the predictions
made actuarially [statistically] were either approxi-
mately equal or superior to those made by a clini-
cian” (p. 119). In a later survey of additional
research, Meehl (1965) reaffirmed his earlier con-
clusions. However, Meehl (1954) also observed
that, in several studies, statistical predictions were
made on the same data from which the regression
equations were developed. In short, the formulas
were not cross-validated. As noted earlier, such for-
mulas frequently show a marked reduction in effi-
ciency when they are applied to samples different
from those used in their derivation.
Sawyer (1966) regarded data collected by inter-
view or observation as clinical data. He viewed
inventory, biographical, or clerically obtained data
as statistical or mechanical. Having considered the
methodological problems and the equivocal results
of the studies he examined, Sawyer concluded that
in combining data the mechanical mode is superior
to the clinical mode. However, he also concluded
that the clinical method is useful in the data collec-
tion process. The clinical method can provide an
assessment of characteristics that would not nor-
mally be assessed by more mechanical techniques
of data collection. But once the data (from what-
ever source) are collected, they can best be com-
bined by statistical approaches.
An example of an individual study comparing
clinical and statistical prediction may help further
illustrate the nature of this controversy. One of the
most frequently cited studies of clinical versus statis-
tical prediction was reported by Goldberg (1965).290 CHAPTER 10