Structured Interview of Personality Organization (STIPO): Preliminary Psychometrics in a Clinical Sample

Abstract

In this article, we describe the development and preliminary psychometric properties of the Structured Interview of Personality Organization (STIPO), a semistructured interview designed for the dimensional assessment of identity, primitive defenses, and reality testing, the three primary content domains in the model of personality health and disorder elaborated by Kernberg (1984; Kernberg & Caligor, 2005). Results of this investigation, conducted in a clinical sample representing a broad range of personality pathology, indicate that identity and primitive defenses as operationalized in the STIPO are internally consistent and that interrater reliability for all 3 content domains is adequate. Validity findings suggest that the assessment of one's sense of self and significant others (Identity) is predictive of measures of positive and negative affect, whereas the maladaptive ways in which the subject uses his or her objects for purposes of regulating one's self experience (Primitive Defenses) is predictive of measures of aggression and personality disorder traits associated with cluster B personality disorders. We discuss implications of these findings in terms of the theory-driven and trait-based assessment of personality pathology.

Acknowledgments

This study was supported in part by grants from the International Psychoanalytic Association to principal investigator B. L. Stern and E. Caligor and a grant from the American Psychoanalytic Association to E. Caligor. We thank Jill Delaney and Jorge Cassab for their contributions to this work and Steven Roose, Andrew Skodol, and Paul Pilkonis for their reviews of an earlier version of this manuscript.

Notes

¹Test–retest reliability, an additional and essential component of reliability testing, is currently being investigated in a separate study.

^a N = 54.

^b N = 88.

²Study participants who reported taking either a mood-stabilizer or anti-psychotic medication were not doing so due to a disclosed bipolar or psychotic condition, which were exclusionary criteria for the study, but rather, we presume, as part of the medical management of other, included conditions.

³The SNAP personality disorder symptom indices are based on DSM–III–R criteria; the SNAP was revised to reflect changes in DSM–IV after data collection for this study began. Dimensional SNAP symptom indexes for DSM–III–R and DSM–IV are highly correlated, with convergence coefficients ranging from .81 to 1.00 and an average of .95 (L. A. Clark, personal communication, September 25, 2005).

*p < .05.

**p < .01.

***p < .001.

⁴Note that the correlations among the independent and dependent variables in Table 3 are presented simply for descriptive purposes and do not represent a series or family of a priori tests.