The Schizotypal Ambivalence Scale: An Item Response Theory Analysis

Abstract

Ambivalence has a prominent role in the historical formulations of schizotypy and schizophrenia, as well as borderline personality disorder. However, it has been overlooked by our current diagnostic nomenclature. The Schizotypal Ambivalence Scale (SAS) is a 19-item self-report scale developed to examine ambivalence relevant to schizotypy and schizophrenia-spectrum disorders. Questionnaire, interview, and ambulatory assessment studies support the construct validity of the measure as a predictor of schizophrenia-spectrum and borderline psychopathology. However, studies have not adequately examined the item properties and factor structure of the scale. To examine the psychometric features of the SAS, the present research applied item response theory and differential item functioning methods using a large sample of adults (n = 7096). Analyses of dimensionality were consistent with essential unidimensionality, and a 2PL IRT model found good item discrimination, an appropriate range of item difficulty, minimal local dependence, and excellent item fit. Analyses of differential item functioning found essentially no bias for gender on any items and very small effects for two items for racial/ethnic identity. Overall, the analyses reveal many psychometric strengths of the Schizotypal Ambivalence Scale and support its use a single-factor instrument for assessing ambivalence in diverse subgroups of adults.