Students in Introductory Psychology (N = 75) were administered the Narcissism-Projective (N-P) which asks subjects to describe two Thematic Apperception Test (TAT) cards and two early childhood memories. Protocols were scored using criteria adapted from the DSM-III description of the narcissistic personality disorder. On the basis of these ratings, the high and low narcissists (n = 16) were chosen for further study. These subjects were interviewed by an experienced clinician who was unaware of their performance on the N-P. Results indicate satisfactory interjudge reliability, p < .05, and clinical validity (the association between N-P and interview ratings of narcissism),/? < .01. It is concluded that, given the N-P's reliability, validity, and psychodynamic significance, this projective measure is useful, generally, as an instrument in the scientific study of psychoanalysis and, specifically, in the rigorous comparative testing of divergent etiological theories of narcissism.The history of psychoanalysis is replete with theoretical controversies that have caused important unresolved rifts. The earliest of these, between Freud and Adler and Freud and Jung, focused on central questions related to human motivation. In order to prove a particular point of view, the advocate would write lengthy theoretical treatises and cite case history data to support the given position. Empirical investigation with the possibility of Requests for reprints should be sent to

This article presents data from a double‐blind experimental study of the comparative validity of specific premises of Kernberg's and Kohut's theories of narcissism. The Narcissistic Personality Inventory (NPI) was administered to 70 undergraduates. Thirteen of the 14 most narcissistic subjects (from the NPI data) participated in three experimental sessions during which the subject was exposed to one of three subliminal stimuli administered tachistoscopically. Each of the experimental stimuli represented an aspect of Kernberg's or Kohut's clinical theory of narcissism. A third stimulus contained neutral content and served as a control. After each tachistoscopic administration, the Narcissism‐Projective (N‐P) and the Self Focus Sentence Completion (SFSC), measures designed to assess narcissism and egocentricity, were administered. Results indicate that the Kernberg stimulus demonstrated a significant effect on the subject's narcissism and egocentricity (p < 0.01). The stimulus that represented Kohut's theory did not demonstrate any statistically significant effect. It is concluded that methodologically rigorous investigations of psychoanalytical questions of theoretical and clinical significance are possible and necessary.

Students in Introductory Psychology (N = 75) were administered the Narcissism‐Projective (N‐P), which asks subjects to describe two TAT cards and two early childhood memories. Protocols were scored using criteria adapted from the DSM‐III section that describes the narcissistic personality disorder. On the basis of these ratings, the high and low narcissists (N = 16) were chosen for further study. These subjects were interviewed by an experienced clinician who was unaware of their performance on the N‐P. The interviewer rated each subject as either a high or low narcissist. After the interview, the subject was administered the Narcissistic Personality inventory (NPI). Results indicate a statistically significant relationship between the interview assessment of narcissism and the N‐P (p < 0.01), the NPI (p < 0.05), and the NPI factor of self‐absorption/self‐admiration (p < 0.05). The limitations and implicaions of these data are discussed. It is concluded that both measures demonstrate promise as instruments useful in the rigorous scientific investigation of significant clinical and theoretical issues related to narcissism.