Reliability and Validity. As with the Rorschach,
let us conclude this section with a summary evalua-
tion of the TAT. It is very difficult to evaluate the
reliability and validity of the TAT in any formal
sense (Lilienfeld et al., 2000). There are so many
variations in instructions, methods of administra-
tion, number of cards used, and type of scoring
system (if any) that hard conclusions are virtually
impossible. The same methodological issues arise
when studying reliability. For example, personality
changes may obscure any conclusions about test–
retest reliability, or there may be uncertainty
about equivalent forms when trying to assess
alternate-forms reliability. It is possible to investi-
gate theme reliability, but since one cannot expect
word-for-word similarity from one occasion to the
next, one is usually studying the reliability of
judges’interpretations. When there is an explicit,
theoretically derived set of scoring instructions
(e.g., B. J. Fitzgerald, 1958), interjudge agreement
can reach acceptable proportions. Interjudge reli-
ability can also be achieved when quantitative rat-
ings are involved (Harrison, 1965). But broad,
global interpretations can present problems.
Some attempts have been made to establish the
validity of the TAT. Methods have included (a)
comparison of TAT interpretations with case data
or with therapist evaluations of the patient; (b)
matching techniques and analyses of protocols
with no additional knowledge about the patient;
(c) comparisons between clinical diagnoses derived
from the TAT and psychiatrists’judgments; and (d)
establishment of the validity of certain general prin-
ciples of interpretation (e.g., the tendency of the
person to identify with the hero of the story or
the probability that unusual themes are more signif-
icant than common ones).
In their review of the research evidence sup-
porting the TAT, Lilienfeld et al. (2000) concluded
that, although there is preliminary, modest support
for the ability of certain TAT scores to assess need for
achievement and for object relations, it is not clear
that the routine use of the TAT is warranted.
Adequate norms for TAT scores are lacking, and
more important, clinicians typically do not compute
scores but rather rely on qualitative impressions of
the stories provided by respondents (Hunsley &
Mash, 2007).
Therefore, typical clinical use of the TAT sug-
gests that it remains basically a subjective instrument.
Although it is possible to identify general principles
of interpretation, these can serve only as guides—not
as exact prescriptions for interpretation. Adequate
interpretation depends on some knowledge of the
patient’s background. As the clinician examines the
test protocol, attention must be paid to the fre-
quency with which thematic elements occur, the
unusualness of stories, the manner in which plots
are developed, misrecognitions, the choice of
words, identifications with plot characters, and so
on. The clinician will want to look closely at the
nature of the TAT heroes or heroines and at their
needs and goals. The situations described by the
respondent are also important, as is the general emo-
tional ambiance of the themes.Sentence Completion Techniques
A very durable and serviceable, yet simple, technique
is the sentence completion method. The most
widely used and best known of the many versions
is the Rotter Incomplete Sentences Blank (Rotter,
1954; Rotter & Rafferty, 1950). TheIncomplete Sen-
tences Blank (ISB)consists of 40 sentence stems—for
example,“Ilike...,”“What annoys me ...,”“I
wish...,”and“Most girls....”Each of the comple-
tions can be scored along a 7-point scale to provide a
general index of adjustment–maladjustment (e.g.,
Jessor, Liverant, & Opochinsky, 1963). The ISB
has great versatility, and scoring techniques for a
variety of variables have been developed (e.g.,
Fitzgerald, 1958). Similar versions of sentence com-
pletion techniques have been developed for use with
young children and with adolescents.
The ISB has several advantages. The scoring is
objective and reliable, due in part to extensive scor-
ing examples provided in the manual. The ISB can
be used easily and economically, and it appears to be
a good screening device. Although it can be scored
objectively, it also allows considerable freedom of
response. Thus, the ISB falls somewhere between
the two extremes of the objective–projective244 CHAPTER 8