measure all important aspects of the construct of
interest? For example, if an interview is designed
to measure depression, then we would expect it
to contain multiple questions assessing various emo-
tional, cognitive, and physiological aspects of
depression.
Criterion-related validityrefers to the ability of a
measure to predict (correlate with) scores on other
relevant measures. These measures may be admin-
istered concurrently with the interview (concurrent
validity) or at some point in the future (predictive
validity). For example, an interview assessing con-
duct disorder in childhood may be said to have
criterion-related validity to the extent that its scores
correlate with measures of peer rejection and
aggressive behavior.
Discriminant validity refers to the interview’s
ability not to correlate with measures that are not
theoretically related to the construct being mea-
sured. For example, there is no theoretical reason
a specific phobia (e.g., of heights) should be corre-
lated with level of intelligence. Therefore, a dem-
onstration that the two measures are not
significantly correlated would indicate the specific
phobia interview’s discriminant validity.
Finally,construct validityis used to refer to all of these
aspects of validity. Thus, many researchers describe
the process of developing and validating a measure as
a process of construct validation. Table 6-6 describes
these validity indices.
In the case of structured diagnostic interviews,
content validity is usually assumed because these
interviews were developed to measure theDSM
criteria for specific mental disorders. That leaves
the need for validation efforts aimed at establishing
an interview’s criterion-related, discriminant, and
construct validity.
Although some validation studies have been
conducted, many more studies are needed. Let us
take a popular structured diagnostic interview, the
SCID, as an example. As noted by Rogers (1995),
relatively few studies have attempted to demonstrate
the criterion-related or discriminant validity of the
SCID. Specifically, not many studies have compared
SCID diagnoses and scores to those obtained from
other diagnostic interviews, to clinical diagnoses, or
to scores from self-report inventories.
Why is this the case? L. N. Robins (1985) has
noted several difficulties associated with validating a
structured diagnostic interview. Laboratory tests that
validate mental disorder diagnoses are not available.
Therefore, there is no“gold standard”to use as a com-
parison. Further, structureddiagnostic interviews were
developed at least partially because of a dissatisfaction
with self-report inventories. Therefore, it does not
seem appropriate or desirable to use a self-report
inventory as a gold standard. Any lack of agreement
between a structured diagnostic interview and a self-
report inventory may be more of an indictment against
the validity of the self-report inventory than it is against
the validity of the structured diagnostic interview. The
same problem exists with using a“clinical impression”
(based on an unstructured clinical interview) as a com-
parison. Some investigators have used a test–retestT A B L E 6-6 Common Types of Validity That Are Assessed to Evaluate Interviews
Type of Validity Definition
Content validity The degree to which interview items adequately measure the various aspects of the variable
or construct.
Predictive validity The degree to which interview scores can predict (correlate with) behavior or test scores that
are observed or obtained at some point in the future.
Concurrent validity The extent to which interview scores are correlated with a related, but independent, set of
test/interview scores or behaviors.
Construct validity The extent to which interview scores are correlated with other measures or behaviors in a
logical and theoretically consistent way. This will involve a demonstration of both convergent
and discriminant validity.
NOTE: Predictive and concurrent validity are both subtypes of criterion-related validity.
THE ASSESSMENT INTERVIEW 187