A Response from the Authors
A Clearer Picture of TAT-Based Measures of Needs

By Scott O. Lilienfeld¹, James M. Wood², and Howard N. Garb³

First, contrary to their claims, our article did not concentrate on the misuse of the Thematic Apperception Test (TAT) or other projective techniques in clinical work, but rather on the question: How well do projective techniques stand up to scientific scrutiny when they are administered and interpreted as intended by their proponents? All psychological measures can be misused, and the prevalence and sources of such misuse are legitimate areas of study. But such misuse was not a major focus of our review.

Second, Woike and McAdams overstate the strength of the evidence for the construct validity of TAT-based measures of needs. For example, Spangler's (1992) meta-analysis of TAT-based measures of achievement motivation revealed that such measures exhibit convergent validity with two classes of real-world achievement measures. But the absolute magnitude of this validity was relatively low, with average rs of .19 and .22. Moreover, findings from studies on TAT-based measures of needs have sometimes proven difficult to replicate (McCrae & Costa, 1984).

As noted in our review, there is relatively little evidence that TAT-based measures of achievement display incremental validity beyond measured intelligence. Moreover, the incremental validity of TAT-based measures of power and affiliation beyond self-report measures of these constructs, although promising (e.g., Woike, 1995), requires further investigation. We agree with Kuncel, Hezlett, and Ones (2001) that "The burden of proof for a new predictor should lie with its proponent, who should demonstrate its incremental validity. This demonstration must take the form of multiple validations across several (large) samples and multiple criterion measures (p. 176)."

Third, we explicitly acknowledged evidence supporting the construct validity of TAT-based measures of achievement, power, and affiliation needs in our review (p. 42). We concluded that TAT-based measures of achievement motivation were among the few projective indexes that satisfied our criteria for empirical support (p. 54), although we noted that the absence of adequate population norms renders their application to clinical settings premature. We concur with Woike and McAdams that TAT-based measures of needs provide a potentially fertile source of research for psychological scientists. We also believe that such measures hold considerably more promise than impressionistic methods of scoring and interpreting the TAT, which are used by the majority of practitioners who administer the TAT (Pinkerman, Haynes, & Keiser, 1993).