ad@dubay.bz
(907) 223 1088
Personality tests, long a mainstay of clinical training, comprise a diminishing portion of the graduate school curriculum despite increasing evidence of their validity. While defensible given the expanded assessment toolkit and other innovations of contemporary clinical practice, this trend has some limitations. For example, the open-ended verbal productions elicited by sentence stems and story prompts closely resemble the material generated by clients in therapy. Mastery of systematic approaches to the coding of sentence completions and TAT stories enables students to develop interpretive skills that will serve them well in standard clinical practice. This paper highlights several rationales to support this conclusion.
The purpose of this commentary is to develop an argument in favor of a renewed focus on performance-based (or projective) methods of personality assessment in clinical training. This proposal is based less on the importance of projective methods in daily clinical practice than on their value as tools for the refinement of our students’ clinical skills. Let me begin with a brief outline of the major problem faced by aficionados of performance-based assessment. Most of my points in this section should be familiar to readers of this journal, but they are rarely integrated into a cohesive framework – and their cumulative impact is, indeed, potentially cataclysmic.
It took half a century for the debate surrounding projective methods to move from extreme viewpoints (“X-Ray of the Psyche” vs. “tea leaves”) to something resembling an acceptable scientific discussion. Any sensible psychologist, whose mind was not closed decades ago, should agree that some performance-based personality assessment methods yield valid information (Mihura et al., 2017; Piotrowski, 2015a); likewise, that there are legitimate grounds for questioning (and investigating) whether these instruments offer adequate incremental validity (after interviews and self-report inventories are administered) to justify the investment of professional time and expertise involved in their use. Given the facts that projective methods correlate weakly with self-reported characteristics and that the two approaches predict distinct (albeit overlapping) behaviors, it is likely that the final scientific verdict will be more favorable than otherwise (Schultheiss et al., 2009). But even as their scientific foundation is strengthened, the place of projective tests in the assessment ecosystem is endangered. Long a central component of training in clinical psychology and related fields (Belter & Piotrowski, 2001), signs of a diminishing status began to emerge a generation ago (Piotrowski & Zalewski, 1993), a trend that has accelerated in the present century (Piotrowski, 2015b; Stedman, Essery, & McGeary, 2018). By the time the science fully demonstrates their value, expertise in their use may be functionally extinct (Piotrowski, 2017).
The declining use of performance-based personality measures by today’s psychologists stems from the convergence of several factors. First, we must recognize the bitter historical feud between tough-minded, nomothetically inclined, statistically savvy academic psychologists and tender-minded, idiographically inclined, relationally focused practitioners. This was an “even fight” for some years, but in the end, the academics (being mostly in charge of training each new generation of practitioners, hence the more “fertile” group) were statistical favorites to emerge victorious. The sluggish initial response of the psychoanalytically fixated but psychometrically challenged practitioners to this challenge did not help matters, either for projective tests or for the psychodynamically informed therapies with which they had become associated. Second, the advent of managed care towards the end of the 20th century forced practitioners to place greater emphasis on short-term, readily measured behavioral change than on in-depth understanding of human individuality (Piotrowski, 1999). It is difficult to justify devoting hours to formal assessment when treatment will be limited to, say, 10 sessions. (The therapeutic assessment movement pioneered by Finn [1996] represents an interesting adaptive response to this challenge.) Third, the emergence of important new methods of assessment (notably semi-structured interviews, clinical rating scales, and behavior checklists) made for a more “crowded” assessment curriculum, and clinical trainers needed to reconsider the place of traditional tests in their syllabi. Fourth, as computer-based test scoring became more widely available and sophisticated, self-reports grew even easier to use relative to labor-intensive projective techniques. This last trend combined with advanced computer-driven forms of data analysis and on-line sampling to make brief self-report measures even more appealing to researchers. (How far this is owing to the need to keep up with an accelerating pace of professional life and how far to a disinclination to engage in effortful mental tasks is uncertain. There are clear advantages to having a billable report largely composed by the ghost in the machine, and when clever multivariate manipulations enable a study based on a few dozen closed-ended items to be published in a mid-ranking personality journal, how many tenure-hungry assistant professors will resist?)
It can be argued that a narrow focus on measures that are easily administered, scored, and analyzed represents an evolutionary dead end for clinical and personality psychology alike. We have long known that the human capacity for reasoned information processing is quite limited relative to that of algorithmic, actuarial decision rules. The end result of our present single-minded devotion to complex statistics will, in all likelihood, be our professional obsolescence. (I wish to emphasize that I am by no means unfriendly either to nomothetic science or statistical analysis – far from it! But I am concerned by our retreat from the attempt to understand people qua people.) Paul Meehl (1973), whose dissertation first confronted us with the superiority of actuarial (vs. clinical) prediction, also wrote presciently about the value of the clinician: as a fellow human being, a psychologist commands a highly evolved capacity to perceive certain qualities in her or his conspecifics, a capacity beyond the ken of even the most powerful computers. Our ability to work with client responses to ambiguous stimuli under open-ended instructions is a case in point.
Still, no matter how great our superiority over computers or poorly-trained technicians in the interpretation of projective material may be, and no matter how strong the evidence grows that performance-based methods yield useful information not readily obtained via other means (see Meyer et al., 2018), the fact remains that the role of performance-based personality assessment will remain relatively circumscribed for the foreseeable future. Given this stark reality, it is reasonable for directors of clinical training programs to ask (as they do!): “why should we devote substantial amounts of precious instructional time to a skill that is popular neither in clinical nor research settings?”
Allow me to introduce my response to that question with an anecdote. When I attended a Boulder-model clinical psychology program, one of the senior professors remarked that today’s psychologists seemed much more skilled at producing change in clients, but much less skilled at decoding their personality patterns: excellent interventionists, but not always understanding the nature of the psychic territory in which they meddled. And this was a quarter-century ago; the trend has accelerated since. Whenever we produce change without understanding the wider system in which our work is embedded, we risk unanticipated complications. Pesticides kill pollinators; antibiotics kill benign symbiotes; antipsychotics trigger metabolic syndrome.
Now, consider the nature of medical training. We do not place a first-year student in an operating room with a hemorrhaging gunshot victim and tell her or him to save the patient. To do so would condemn said patient to inept care, even iatrogenic injury. We begin with laboratory exercises in anatomy and physiology; dissection of infrahuman animals giving way in time to detailed work with cadavers. In this way, students acquire familiarity with the various anatomical and physiological processes, quick to recognize what they are dealing with and confident in their ability to navigate the complex internal space of the body.
No more should we place beginning trainees in counseling and psychotherapy, whose background to date is all based on textbooks and lectures, in settings where they must identify and cope with the psychic traumas of living, breathing patients. The internal space of the mind is no less complex than that of the physical body, and just as easily damaged by a clumsy intervention. Yes, we have students practice interviewing skills with one another and observe themselves on tape, but this interpersonal sparring (valuable as it may be!) is only a shadow of the reality that awaits in the clinic: the live-action engagement with conflicted and suffering humanity. We want our graduate students to learn how to perceive and interpret the meanings of client utterances – to recognize the nuances of individual statements – to discern the underlying narrative structure of memories, fantasies, and (perchance) dreams. Where are we to find the preserved remains of complex psychic phenomena? What is our equivalent of the frog, the fetal pig, the cadaver?
I suggest that certain forms of performance-based (or projective) material are ideally suited for this purpose. In particular, I would like to call for a renewed emphasis, in clinical training and research alike, on two projective methods: sentence completions and the picture-story exercise exemplified by the Thematic Apperception Test (TAT). These provide us with samples of verbal behavior, extracted from actual people under controlled conditions and preserved intact for our continued examination and study.
Consider, first, sentence completion tests. Here, “stems” may be constructed ranging from highly ambiguous (e.g., “I think…”) to highly structured (e.g., “When a woman is put in charge…”) depending on the focus of the test designer’s interest. The most open-ended stems will elicit maximally varied responses and allow for flexible scoring and interpretation, while more specific prompts will “pull” for narrowly defined response categories, useful when one is investigating a particular construct. Each response can be scored independently. Since the responses are succinct, a relatively large number of items can be included. These factors enhance reliability. But from a clinical training perspective, each response represents a distinct personal statement of the sort that clients in therapy often make. As clinical students learn the possible meanings of the many responses offered up that begin with, say, “I think,” they are learning how to “read into” the statements or clinical material their future clients may actually provide in treatment.
We already possess reliable, well-validated scoring systems enabling us to extract a wide variety of meanings from sentence completions. The most widely used clinical test of this type, the Rotter Incomplete Sentences Blank (RISB; Rotter, Lah, & Rafferty, 1992), features a well-articulated system whereby we may evaluate the level of Adjustment (vs. Maladjustment) exhibited by any given response. The system balances positive responses against conflictual ones, so that strengths and vulnerabilities both may be incorporated into our case formulations. Also, because of the inherent flexibility of the method, it is possible to develop additional scoring systems that can be applied to the same test protocols. To take just three examples, it is possible to use RISB protocols to assess defense mechanisms (Johnson & Gold, 1995), cognitive distortions (Lehnert, Overholser, & Adams, 1996), and major personality traits (Joy, 2017).
Less frequently seen in clinical settings, but the subject of an even more impressive body of research, is Loevinger’s Washington University Sentence Completion Test (WUSCT; Hy & Loevinger, 1996), which rejoices in an elaborate, well-validated system for evaluating one’s level of ego development, from primitive impulsive characters up to highly integrated and differentiated identities. Aside from these two measures, most of the remaining sentence completion tests are not well known and have, in many cases, fallen largely into disuse (Rohde [1947], used to evaluate motives, is a good example) –though there have been encouraging signs of interest in developing specialized contemporary measures of this type to assess depression (Barton et al., 2005), vulnerability to bipolar disorder (Dempsey, Gooding, & Jones, 2014), psychopathy (Endres, 2004), and poor performance in military training (Picano et al., 2002).
My point is that in training students to discern the significances underlying sentence completions, we are not merely training them to score more-or-less obscure personality tests. Rather, we are refining their clinical sensibilities, teaching them to master the skill of finding meaning in the many personal statements made by clients during intake interviews and subsequent therapeutic sessions. And we are doing so in an entirely safe environment. The statements lie before us as on the anatomist’s dissection table, there to be studied at leisure and shorn of any risk. Misinterpretations will harm nobody as the professor circulates among the students, gently guiding their perceptions. We may progress from analysis of well-studied protocols supplied by the professor (psychological autopsies, as it were) to case material obtained by the students. With practice, the students will become swift and sure, able to intuit meaning in real time during clinical interviews.
As the sentence completion method is mastered, attention may be shifted onto the more complex TAT-type measures. Here, the individual samples of verbal behavior are much larger and the challenge of extracting meaning in a principled, reliable fashion much greater. But the potential rewards are concomitant with the effort required. For what we are presented with are complete narratives: representations of the social cognitive structure of the respondent’s mind. Murray’s original idea may have been that fantasy can substitute for dream material as a way of exploring the psyche, but the picture-story method also parallels episodic memory, especially those episodic memories found most important by the reflective, self-creating human mind. Memories, dreams, waking fantasy, anticipated futures: these are the stuff of which we are made.
I do not think most clinicians – or most researchers – fully appreciate the extent to which the entire narrative approach to personality derives from the use of TAT-type measures. Nor are most aware of the range of uses to which such measures have been put. We tend to tell a story that begins (appropriately) with Murray’s (1938) “needs and press” model, maybe including some remarks along lines of how the interaction of needs and press, leading to a given outcome, constitutes a simple thema. We then consider either the purely clinical tradition of TAT use (e.g., Bellak, 1996) or its applications in pure research (as, for example, in McClelland’s classic studies of power and achievement motivation). If we are a bit more sophisticated, we consider both traditions and lament the lack of meaningful bridges uniting the two. There is nothing incorrect about such an approach, but it fails to do justice to the history of testing.
Variations on the picture-story exercise have been used in cross-cultural studies by anthropologists (including classic work conducted in Japan, in Latin America, and among southwestern Native American tribes). Scoring systems have been validated not only for Murray-style needs, motives, or central concerns (Smith, 1992), but also for primary process thinking (Holt, 2008), defense mechanisms (Cramer, 1991), and social cognition/object relations (Westen, 1995). The failure of clinical psychology faculty to teach such approaches (many of which are reviewed by Jenkins, 2015) contributed much to the method’s gradual decline in popularity.
But there is more. It should be remembered that Silvan Tomkins, whose script theory (1995; Sedgwick & Frank, 1995) forms the foundation for today’s narrative models of personality, began as one of Murray’s research associates. His analysis of personality in terms of nuclear episodes, scenes that slowly transmute into the scripts by which we live, stems directly from his earlier work analyzing TAT stories. Without Tomkins we would not have the contemporary researchers (e.g., McAdams, 2016) who have explored the narrative structure of self: identity as constructed on the basis of affectively charged self-defining memories. For any approach taken to the study of TAT stories can also be applied to other narrative forms: dreams, yes, but also recollections of early childhood or simple accounts of last week’s family outing.
In short: as clinical students master the craft of decoding the many layers of meaning inherent in TAT protocols (or any similar narrative samples, such as early memories), these future practitioners are becoming more adept at discerning the abstract essence of people’s stories. And stories are the working material of psychotherapy as of human life more generally. (This work, conducted in assessment classes and workshops, would unfold in parallel with the cultivation of interviewing skills in psychotherapy methods courses.) It may not much matter whether they will make extensive use of the TAT or its cousins in their later clinical practice. What matters is that by studying these stories, duly recorded and safely ensconced in the laboratory, they are developing crucial clinical skills. Grotesque as the metaphor may appear: if sentence completions are the fetal pigs of personology, fantasies and memories are our cadavers.
Even if nothing new were ever to be discovered using these methods, any more than new discoveries are likely when well-known creatures are placed on the dissection table, their use in clinical training would be more than justified. But I will go further, and assert that it is through the study of people’s free responses to structured but open-ended prompts that we are likely to make new findings. Closed-ended questionnaires, for all their very real value, operate as closed systems, enabling us to test well-formed hypotheses but never introducing anything unexpected. Neither object relations theory nor the narrative approach to personality could possibly have emerged from a generation of psychologists schooled only in the statistical analysis of self-report data. If we want the next generation to be capable of fresh insights into human nature, we need to give them every opportunity to explore the ways in which our inner worlds are revealed by our expressive behaviors. Training in projective methods such as the use of sentence completion and storytelling tasks is one of the principal means by which this important aspect of our mental life may be achieved.
References:
Barton, S., Morley, S., Bloxham, G., Kitson, C., & Platts,
Bellak, L., & Abrams, D.M. (1996). The T.A.T., the C.A.T., and the S.A.T. in clinical use (6th ed.). Boston, MA: Allyn & Bacon.
Belter, R.W ., & Piotrowski, C. (2001). Current status of doctoral-level training in psychological testing. Journal of Clinical Psychology, 57(6), 717-726.
Cramer, P. (1991). The development of defense mechanisms: Theory, research, and assessment. New York: Springer-Verlag.
Dempsey, R.C., Gooding, P.A., & Jones, S.H. (2014). Assessing the specificity of autobiographical memory in individuals at a trait-based vulnerability to bipolar disorder using a sentence completion task. Memory, 22, 222-231.
Endres, J. (2004). The language of the psychopath: Characteristics of prisoners’ performance on a sentence completion task. Criminal Behaviour and Mental Health, 14, 214-226.
Finn, S.E. (1996). Manual for using the MMPI-2 as a therapeutic intervention. Minneapolis, MN: University of Minnesota Press.
Holt, R.R. (2008). Primary process thinking: Theory, measurement, and research. Lanham, MD: Jason Aronson.
Hy, L.X., & Loevinger, J. (1996). Measuring ego development (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates.
Jenkins, S.R. (Ed.). (2015). A handbook of clinical scoring systems for Thematic Apperception Techniques. New York: Routledge.
Johnson, N.L., & Gold, S.N. (1995). The Defense Mechanism Profile: A sentence completion test. In H.
Joy, S.P. (2017). Reading personality: Assessing “Big Three” traits with the sentence completion method. Current Issues in Personality Psychology, 5, 215-231. Lehnert, K.L., Overholser, J.C., & Adams, D.M. (1996). The Cognition Rating Form: A new approach to assessing self-generated cognitions in adolescent sentence completions. Psychological Assessment, 8,
172-181.
McAdams, D. (2016). The art and science of personality development. New York: Guilford Press.
McClelland, D.C., Koestner, R., & Weinberger, J. (1989). How do self-attributed and implicit motives differ? Psychological Review, 96, 690-702.
Meehl, P. (1973). Psychodiagnostics: Selected papers.
Minneapolis, MN: University of Minnesota Press.
Meyer, G.J., Katko, N.J., Mihura, J.L., Klag, M.J., & Meoni, L.A. (2018). The incremental validity of self-report and performance-based methods for assessing hostility to predict cardiovascular disease in physicians. Journal of Personality Assessment, 100, 68-83.
Mihura, J.L., Roy, M., & Graceffo, R.A. (2017). Psychological assessment training in clinical psychology doctoral programs. Journal of Personality Assessment, 99, 153-164.
Murray, H. (1938). Explorations in personality. New York: Oxford University Press.
Picano, J.J., Roland, R.R., Rollins, K.D., & Williams, T.J. (2002). Development and validation of a sentence
completion measure of defensive responding in military personnel assessed for nonroutine missions. Military Psychology, 14, 279-298.
Piotrowski, C. (2017). The linchpin on the future of projective techniques: The precarious status of personality assessment in the (overcrowded) professional psychology curriculum. Journal of Projective Psychology and Mental Health, 24, 71-73.
Piotrowski, C. (2015a). Projective techniques usage worldwide: A review of applied settings 1995-2015. Journal of the Indian Academy of Applied Psychology, 41(3), 9-19.
Piotrowski, C. (2015b). On the decline of projective techniques in professional psychology training. North American Journal of Psychology, 17, 259-266.
Piotrowski, C. (1999). Assessment practices in the era of managed care: Current status and future directions. Journal of Clinical Psychology, 55(7), 787-796.
Piotrowski, C.,& Zalewski, C. (1993). Training in psychodiagnostic testing in APA-Approved PsyD and PhD clinical psychology programs. Journal of Personality Assessment, 61, 394-405.
Rohde, A.R. (1947). Sentence Completions Test Manual.
New York: Author.
Rotter, J.B., Lah, M.I., & Rafferty, J.E. (1992). Rotter Incomplete Sentences Blank Manual (2nd ed.). San Antonio, TX: The Psychological Corporation.
Schultheiss, O. C., Yankova, D., Dirlikov, B., & Schad, D.
Sedgwick, E.K., & Frank, A. (Eds.). (1995). Shame and its sisters: A Silvan Tomkins reader. Durham, NC: Duke University Press.
Smith, C.P. (Ed.). (2008). Motivation and personality: Handbook of thematic content analysis. Cambridge, UK: Cambridge University Press.
Stedman, J.M., Essery, J., & McGeary, C.A. (2018). Projective personality assessment: Evidence for a decline in training emphasis. Journal of Projective Psychology and Mental Health, 25, 54-59.
Tomkins, S.S. (1995). Exploring affect: Selected writings (Edited by E.V. Demos). New York: Cambridge University Press.
Westen, D. (1995). Social Cognition and Object Relations Scale: Q-sort for projective stories (SCORS-Q). Unpublished manuscript.
We gratefully acknowledge the support of our sponsors.
© 2026 Somatic Inkblots. All Rights Reserved.