A construct validity study of a Differentiation of Self Scale.

The purpose of this study was to establish construct validity of the Haber Level of Differentiation of Self Scale (LDSS) that measures selected aspects of the construct, differentiation of self, the key concept of the Bowen Theory. During stage one of this study, a 32-item Likert-type LDSS, which contained two subscales, Emotional Maturity, (EM), and Emotional Dependency (ED), was administered to 372 volunteer subjects. Factor analytic and correlational findings indicated support for a 24-item unidimensional rather than a 32-item two subscale instrument. Recalculation of coefficient alpha (.90) also supported the unidimensional structure of the LDSS. During stage two of this study, the revised 24-item LDSS was administered to 372 volunteer subjects, who also completed the State-Trait Anxiety Inventory, the Life Experience Survey, and the Behavior Checklist. Hypotheses predicted that there would be negative relationships between differentiation of self and trait anxiety, state anxiety, adult psychological dysfunction, and negative impact stress. Findings indicate support for all hypotheses at the p < .01 level of significance or greater. Data from this study suggest that the 24-item LDSS is a valid and reliable unidimensional family assessment tool that accurately measures specific aspects of the construct, differentiation of self. The LDSS can be used by nurses and other professionals to assess individual and family functioning in a variety of health care settings.