- Response interpretation deals withmore vari-
ables. Since the range of possible responses is so
broad, the clinician can make interpretations
along multiple dimensions (needs, adjustment,
diagnostic category, ego defenses, and so on).
Many objective tests, in contrast, provide but a
single score (such as degree of psychological
distress) or scores on a fixed number of
dimensions or scales.
Measurement and Standardization
The contrasts between objective tests and projective
tests are striking. The former, by their very nature,
lend themselves to an actuarial interpretive approach.
Norms, reliability, and even validity seem easier to
manage. The projectives, by their very nature, seem
to resist psychometric evaluation. Indeed, some clini-
cians reject even the suggestion that a test such as the
Rorschach should be subjected to the indignities of
psychometrics; they would see this as an assault upon
their intuitive art. In this section, we offer several gen-
eral observations about the difficulties involved in eval-
uating the psychometric properties of projective tests.
Standardization. Should projective techniques be
standardized? There are surely many reasons for doing
so. Such standardization would facilitate communica-
tion and would also serve as a check against the biases
and the interpretive zeal of some clinicians. Further-
more, the enthusiastic proponents of projectives usually
act as if they have norms (implicit though these may be)
so that there seems to be no good reason not to
attempt the standardization of those norms. Of course,
research problems with projectives can be formidable.
The dissenters argue that interpretations from
projectives cannot be standardized. Every person is
unique, and any normative descriptions will inevita-
bly be misleading. There are so many interacting
variables that standardized interpretive approaches
would surely destroy the holistic nature of projective
tests. After all, they say, interpretation is an“art.”
Reliability. Even the determination of reliability
turns out not to be simple. For example, it is surely
too much to expect an individual to produce, word
for word, exactly the same TAT story on two dif-
ferent occasions. Yet how many differences
between two stories are permissible? Of course,
one can bypass test responses altogether and deal
only with the reliability of the personality interpre-
tations made by clinicians. However, this may con-
found the reliability of the test with the reliability of
the judge. Also, test–retest reliability may be
affected by psychological changes in the individual,
particularly when dealing with patient populations.
It is true that clinicians can opt for establishing reli-
ability through the use of alternate forms. How-
ever, how do they decide that alternate forms for
TAT cards or inkblots are equivalent? Even split-
half reliability is difficult to ascertain because of the
difficulty of demonstrating the equivalence of the
two halves of each test.Validity. Because projectives have been used for
such a multiplicity of purposes, there is little point
in asking general questions: Is the TAT valid? Is the
Rorschach a good personality test? The questions
must be more specific: Does the TAT predict
aggression in situation A? Does score X from the
Rorschach correlate with clinical judgments of
anxiety?
With these issues in mind, we turn now to a
discussion of several of the more popular projective
tests.The Rorschach
Although the origins of theRorschachlie in Europe,
its subsequent development and elaboration
occurred in the United States (Butcher, 2010;
Exner, 1993). Disenchantment with objective
inventories probably facilitated this development
(Shneidman, 1965). However, the general rise of
the psychodynamic, psychoanalytic movement and
the emigration of many of its adherents from
Europe to the United States in the 1930s were also
important.
What has confused many and perhaps impeded
efforts to demonstrate reliability and validity is the
fact that there are several different general Rorschach
approaches. For example, in the past, Klopfer, Beck,PERSONALITY ASSESSMENT 237