Not Your Same Old Story: New Rules for Thematic Apperceptive Techniques (TATs)

ABSTRACT

Stories told about pictures have been used for both research and clinical practice since the beginning of modern personality assessment. However, with the growing science–practice gap, these thematic apperceptive techniques (TATs) have been used differently in those 2 venues. Scientific validation is presumptively general, but clinical application is idiographic and situation-specific. A bridge is needed. The manualized human-scored narrative analysis systems discussed here are valuable scientist-practitioner tools, but they require a validation literature to support further research publication, maintain their role in clinical training, and justify clinicians' reimbursement by third-party payers. To facilitate wider understanding of manualized TAT methodologies, this article addresses long-standing criticisms of TAT reliability and proposes some strategic solutions to the measurement error problem for both researchers and clinicians, including analyzing person–situation interactions, purposeful situation sampling for within-storyteller comparisons, and uses of small samples. The new rules for TATs include conceptual and methodological standards that researchers should aim to meet and report, reviewers should apply to manuscripts, and clinical assessors can use to analyze their own data and justify third-party payment.

Acknowledgments

I thank William D. Spangler, two anonymous reviewers, Rachel B. Nowlin, and Molly S. Tucker for their comments on previous versions.

Notes

¹ This article, intended for this issue's Special Section, was inadvertently published in JPA Volume 98, Number 6. The correct reference is Siefert, C. J., Stein, M. B., Slavin-Mulford, J., Sinclair, S. J., Haggerty, G., & Blais, M. A. (2016). Estimating the effects of Thematic Apperception Test card content on SCORS–G ratings: Replication with a nonclinical sample. Journal of Personality Assessment, 98, 598–607.

² In describing the measurement model in which items define the construct (Streiner's [2003] index), Bollen and Lennox (1991) insisted that conceptualizing items as “sampling” the facets of the construct is hazardous because “omitting an indicator is omitting a part of the construct … we need a census of indicators, not a sample” (p. 308). I propose here a composite conceptualization based on the stimulus sampling frame, in which each cell of the frame represents an identified facet, within which cards sharing those characteristics are seen as functionally equivalent indicators, from which a sample can be drawn.

³ Predicting aggressive behavior has been a long-standing challenge for all areas of psychology, not only clinical assessment, because of the complexity of the pathway from mental states to behavior. One excellent TAT study in this line was by Minuchin (1965), whose TAT scoring for aggression and various “modifying mechanisms” (defenses and aggression anxiety) predicted behavioral aggression among hospitalized neuropsychiatric patients.

⁴ The use of undergraduate samples in the psychological literature rests on an assumption of generalizability that is long known to be false on the rare occasions tested, but that is concealed by collective amnesia because it functions as a dirty secret among academic psychologists: For many it is the only way they can get original research done at all, given the shrinking pie of funding from all sources. Recruiting and compensating community-based and patient samples for any purpose is even more resource-intensive than scoring TATs.