Psychometric Properties of Object Relations Scales for the Rotter Incomplete Sentences Blank in an Inpatient Sample.

Object relations (OR) scales were recently developed for the Rotter Incomplete Sentences Blank by Rucker and Krishnamurthy (2023). This study examined the psychometric properties of these scales with N = 160 psychiatric inpatients by evaluating the factor structure, internal reliability, and criterion validity with instruments measuring object relations features. A confirmatory factor analysis indicated a two-factor model was a better fit than a one-factor model; however, neither were excellent fits. Interrater reliability was excellent for the Self-Representation (r = .95), Other-Representation (r = .96), and Total OR scales (r = .97). Internal consistency reliability for the Total OR scale was acceptable at α = .72, but Self-Representation (α = .67) and Other-Representation (α = .58) subscales were in the clinically suboptimal range. Construct validity analyses demonstrated several conceptually relevant convergences with scales from the Inventory for Personality Organization and Bell Object Relations and Reality Testing Inventory, but unconvincing divergences. Group differences in RISB OR scores were observed between inpatients with and without a personality disorder. Altogether, these findings provide mixed psychometric support for the RISB OR scales and raise doubts about their use as standalone instruments. Results suggest a need for continued reevaluation, if not refinement.