The Assessment of the Social Cognition and Object Relations Scale on TAT and Interview Data

Abstract

This study examines the reliability and convergent validity of 2 versions of the Social Cognition and Object Relations Scale (SCORS), one for use with Thematic Apperception Test narratives (SCORS–TAT; Westen, 1990) and one for use with clinical interview data (SCORS–CDI; Westen, Barends, Leigh, Mendel, & Silbert, 1990). Four SCORS dimensions were evaluated. Data were collected in a psychiatric sample (N = 74). Results show that although interrater reliability was good for all dimensions, internal consistency was low, especially for the affective dimensions. Structural equation modeling, in which a model with 2 factors (i.e., SCORS–TAT and SCORS–CDI) and 4 dimensions each was tested, indicated low convergence between corresponding dimensions of SCORS–TAT and SCORS–CDI. Correlational analyses suggested that this was due to a strong method factor. Regression analyses, however, revealed that the presence of a personality disorder operated as a moderator for convergence between corresponding cognitive-structural dimensions.