criterion keying approach, which works well for
those who understand it. But what about people
who know nothing of criterion keying or psycho-
metrics but want to work for Corporation X? Sup-
pose these individuals fail to get a job after taking an
MMPI-2. What shall we tell them when they
demand to know what items such as“I used to
keep a diary”or“My sex life is satisfactory”have
to do with hiring? For the person seeking therapy, a
test such as the MMPI-2 may be acceptable. How-
ever, for one who takes the test in a personnel set-
ting, it may be regarded as an invasion of privacy
(Butcher, 1971).
In fact, given recent court rulings as well as the
requirements specified by the Americans with Dis-
abilities Act, it may not be advisable to use popular
personality/psychopathology tests like the MMPI-2
in most preemployment screening situations
(Camara & Merenda, 2000). Tests that contain
items that inquire about religious beliefs, political
orientation, or sexual orientation have been ruled
violations of privacy, and courts have ruled that
applicants for jobs can only be screened for mental
disabilities once a preliminary job offer has been
made (Camara & Merenda, 2000). In this latter
context, psychological tests are viewed as a type of
“medical examination,”and if a mental disability is
uncovered, it is up to the employer to demonstrate
how such a disability directly and adversely affects
performance of the job in question.
Given the nature of the original sample that the
MMPI was validated on, questions have also been
raised as to whether the instrument may be“biased”
against certain ethnic and racial groups. This has been a
rather contentious debate over the years (e.g., Gynther,
1972; Gynther & Green, 1980; Pritchard &
Rosenblatt, 1980). Some studies have found significant
differences in scores between racial groups, whereas
others have not. It is important to keep in mind,
however, that a significant difference between mean
scores for groups of people doesnot,inandofitself,
indicate test bias. Rather,test biasrefers to a situation
in which different decisions or predictions are made
formembersoftwogroupsevenwhentheyobtain
the same score (Anastasi, 1988). We will return
to the general issue of test bias later in this chapter.
To our knowledge, there have been only a relatively
small number of published studies that evaluate the
possibility of test bias for the MMPI-2. It is encouraging
that, to date, these few studies have not found evidence
suggesting that the use of the MMPI-2 for certain pur-
poses results in bias against certain ethnic or racial
groups (e.g., Arbisi, Ben-Porath, & McNulty,
2002; Timbrook & Graham, 1994; Wood et al.,
2002).Concerns About the MMPI-2. The reasons that
prompted the revision of the MMPI were, in the
main, laudable. Without doubt, though, some clin-
icians were nervous about the MMPI-2 revision.
Some of these concerns have diminished as clinical
psychologists became more familiar with the revi-
sion. Nevertheless, here is a sample of the com-
plaints and reservations that have been voiced
about the MMPI-2.- The normative sample, compared to U.S.
Census Bureau data, is too highly educated.
Only 5% of MMPI-2 normative respondents
have less than a college education, and 45% of
the normal respondents in the sample are college
educated! Schooling can account for much of
the variance in the scores of psychiatric patients. - Is all the old MMPI research applicable to the
new MMPI-2? - Were“unnecessary”revisions made?
- Criteria for the inclusion of“normal”respon-
dents are puzzling. - Some respondents who are administered both
versions of the test show psychological prob-
lems on one version but not on the other. - Scores are generally lower on the MMPI-2
compared to the MMPI. - Validity of the new content scales is unclear.
- It is not clear how reliable MMPI-2 scores are
over extended periods of time. - Internal consistency of several MMPI-2 scales is
quite low. - There remains too much item overlap among
the scales, making study results hard to interpret.
PERSONALITY ASSESSMENT 231