Understanding Psychological Testing
To print a pdf
version of this article, click here
Non-psychologists are often unnecessarily intimidated by
psychological testing. They believe that psychologists can administer
tests that will provide a veritable window into their unconscious and
reveal all their darkest secrets. Test results are too often accepted
uncritically as being 100% accurate, when in fact, all too often there
are serious errors that can impeach the test’s conclusions.
The usefulness of any
psychological test in the forensic arena is largely dependent on two
things: reliability and validity. A test is considered reliable
if test results from the same person remain consistent (within a small
range of error) over time. If a test wasn’t reliable, we wouldn’t know
which result was in fact accurate: was it the result from the 1st
administration of the test or was it the result from the 2nd
administration? Or neither? Now, this presumes that we’re talking
about a relatively short period of time (except for IQ tests, where the
score should remain in the same general range throughout one’s
lifetime). If, for instance, you take the MMPI in August, we’d hope to
get the same general results
as if you took it in September. The more consistent a test is, the more
reliable it is considered to be.
But reliability isn’t worth squat until we look at the test’s
validity. Validity refers to the degree to which a test
measures what it purports to measure. For instance, an IQ test is valid
for measuring intelligence---that is, it measures the subject’s level of
intelligence, but it is not valid for assessing someone’s
How do we know that a test is valid? A commercially-sold test
(which are the only kind that should be used forensically) should have
reliability and validity information as part of the test manual.
(Beware! There are some so-called “tests” that do not have this
data.) Most commercially-sold tests are investigated intensely and the
results used to establish test norms, i.e. statistics that show
how people similar to the person being currently evaluated performed on
this test. The normative group should be identical on all factors of
importance (which will vary from case to case) as the client being
evaluated. For instance, there are specific norms for child
custody litigants on the MMPI. If, for example, a custody litigant’s
MMPI scores are compared to the general population (and not to child
custody litigant norms), the results will not be valid.
The concept of validity is far more complicated that can be
discussed in this brief paper. Suffice it to day, there are many
different components that make up test validity and how it is measured.
Validity and reliability aside, there are different kinds of
psych-ological tests. There are intelligence tests, which measure
someone’s intellectual ability or IQ. There are also innumerable kinds
of personality tests. Many, like the Minnesota Multiphasic Personality
Inventory (MMPI) are “self-report” tests, in which the subject simply
reports his responses, often in a true/false format. There are also
projective tests, which purport to assess some aspect of a subject’s
personality by the way s/he responds to an ambiguous stimulus. The
Rorschach Ink Blots are perhaps the best known (but certainly not the
only) example of a projective test. The subject is shown an ambiguous
stimulus---in this case, the ink blot---and inferences about his
personality are made from what s/he says s/he imagines it might be
(a.k.a. what s/he “projects”). Since the blot is, in fact, nothing at
all, the theory is that what the subject chooses to “see” there (a
process known as “projection”) says something about his or her
There is disagreement amongst psychologists as to whether
or not the entire “projective hypothesis” is even true! And, as you
may imagine, it is considerably more difficult to establish reliability
and validity data for projective tests, and for this and other reasons,
their use in a forensic arena is highly controversial.
Not every forensic mental health question necessitates the
use of testing, by the way, and it is also true that testing potentially
can be challenged by the astute attorney who has taken the time to
understand its nuances. At the same time, appropriate testing can
frequently shed light on previously poorly-understood behaviors.
When should you consider requesting psychological testing?
Whenever your client’s mental health is part of the legal case, testing
is one option that should at least be considered. Here are some
questions to ask the psychologist before authorizing a test battery:
testing needed in this case?
tests do you propose to use?
the reliability and validity statistics?
closely does my client resemble the normative group?
grounds could this test be challenged by opposing counsel?
Testing can be a valuable adjunct to any litigation. But it
is also a potential minefield for the attorney who is unfamiliar with
its specific caveats and limitations. It can provide valuable,
empirically established support for your position, or it can torpedo
the best-laid strategy. I am available to help you critically examine
the role of psychological testing in your case. And, as always, the
first half hour of any consultation is absolutely gratis! Call me at