by MMPI-2 items may suggest certain personality
traits or styles, it was not developed with person-
ality constructs in mind. Does it help with personality
description, psychotherapy planning, or the host of
other predictions that the clinician must confront in
day-to-day interactions with the patient? Many
would question the MMPI-2’s contribution to ongo-
ing clinical decisions once the initial diagnostic cate-
gory has been selected. However, Butcher (1990,
1995c, 2010) would argue that the MMPI-2 is a
valuable aid in planning and evaluating the effects of
treatment.
Reliability and Validity. As noted earlier, there
have been more than 19,000 published studies and
books concerning the MMPI family of measures.
Many of these publications have evaluated the reli-
ability and validity of MMPI-2 scores. Aside from
the comments already made, we do not have the
temerity to “briefly”summarize this voluminous
research. However, we do wish to make a few
comments about findings addressing the reliability
and validity of this instrument’s scores.
Concerning reliability, the MMPI-2 clinical
scale scores have been evaluated in terms of internal
consistency as well as test–retest reliability. Perhaps
not surprisingly (given the way the instrument was
originally constructed), many of the clinical scales
do not have good internal consistency. Only three
of the ten clinical scales (scales 7, 8, and 0) show
good internal consistency across both men and
women (defined as an alpha of .80 or above)
(Butcher et al., 2001). Recall that items for scales
were selected primarily for their ability to discrimi-
nate between groups, and not as a function of their
ability to measure a unitary dimension or construct.
In contrast to internal consistency results, the
MMPI-2 clinical scale scores do show good test–
retest reliability over time, in this case a one-week
interval (Butcher et al., 2001).
The evaluation of the MMPI-2’s validity is
more complicated. As is true with all measures,
we must evaluate the validity of scores for certain
prediction tasks. The ultimate question, then, is not
whether the MMPI-2 is valid. Rather, for what
specific purposes are its scores valid? MMPI-2 scores
have been shown to relate significantly to relevant
external correlates such as emotional states, antiso-
cial behaviors, stress reactivity, worry, paranoia, and
introversion, for example, across a number of popu-
lations and settings (Graham, 2006). Furthermore,
MMPI-2 profiles (patterns of score elevations) also
appear to show validity in identifying characteristic
moods, cognitions, and behaviors of patients who
produce respective“code types”(Graham, 2006).
Two other issues related to the validity of
MMPI-2 scores are important as well. Although
presented in the context of developing “new”
MMPI-2 supplementary scales, Butcher et al.
(1995) noted the necessity of establishing the incre-
mental validity of a scale as well as the validity of
cutoff scores (thresholds).Incremental validityis sup-
ported if scale scores provide information about a
person’s behavior, personality features, or psycho-
pathology features that is not provided by other
measures. Do MMPI-2 scores convey information
relevant to psychopathology or personality that
cannot be provided by other measures? This
issue of incremental validity tends to be neglected
for all psychological tests, including the MMPI-2
(Hunsley & Mash, 2007).
As for thevalidity of cutoff scores, it is important
to keep in mind that the optimal cutoff score (in
terms of maximizing correct decisions as to which
patients have the disorder or trait in question, given
their scores on the measure) will vary depending on
the nature of the population of patients sampled.
The cutoff scores provided in test manuals were
derived for a certain population, which may or
may not be similar to the population of patients
with which a clinician is working. MMPI-2 cutoffs
(Tscore of 65 or greater) were derived using the
distribution of scores from the normative sample.
Therefore, these cutoffs may or may not be appro-
priate in certain clinical contexts.Personnel Selection and Bias. Recent develop-
ments in our society have also created problems for
the MMPI-2. Lack of trust in our social institutions
and the concerns of minorities have both been
reflected in criticisms of the test. For example, the
MMPI-2 has often been lauded for its empirical230 CHAPTER 8