The factor structure of the short form of the Wisconsin schizotypy scales. (July 2018)

• Record Type: Journal Article
• Authors: Preti, Antonio; Scanu, Rosanna; Muratore, Tamara; Claudetti, Giustino; Cao, Andrea; Scerman, Rossana; Carrus, Marta; Cadoni, Carlotta; Manca, Antonio; D'Errico, Giovanni; Contu, Augusto; Petretto, Donatella R.
• Is Part Of: Psychiatry research. Volume 265(2018)
• Page Start: 128
• Page End: 136
• Publication Date: 2018-07
• Subjects: Schizotypy -- Wisconsin schizotypy scales -- Schizophrenia -- Psychosis proneness -- Confirmatory factor analysis; Psychiatry -- Periodicals; Psychiatry -- periodicals; Psychiatrie -- Périodiques; 616.89
• DOI: 10.1016/j.psychres.2018.04.041
• Languages: English
• ISSNs: 0165-1781

Abstract

Highlights: The short-form Wisconsin schizotypy scales (WSS) were investigated. The short-form WSS showed good reliability. Structural validity of the short-form WSS was also provided. A multidimensional model of schizotypy was confirmed. Measurement models should be considered with caution. Abstract: The Chapman psychosis-proneness scales—also known as Wisconsin schizotypy scales (WSS)—are among the most used tools to measure schizotypy. The factor structure of the short-form WSS was investigated in a mixed sample of patients with chronic mental disorders and of healthy subjects from the general population. One hundred patients with a chronic mental disorder were enrolled over a 6-month period. For each patient, two controls of same sex and similar age (±5 years) were enrolled; 131 accepted to take part in the study. The unidimensional, the correlated four-factor, the second-order two-factor models, and the bifactor model with two or four orthogonally independent factors of the short-form WSS were tested with confirmatory factor analysis. Good reliability of the short-form WSS was confirmed, as its capacity of differentiating people with and without schizotypy. The bifactor models were superior to other models. However, in both bifactor models the explained common variance (ECV) attributable to the general factor and the percentage of uncontaminated correlations (PUC) were too low to use a general summary score as a measure of a single latent schizotypy variable. Symptoms scores derived from the short-form WSS can be better appreciated within a multidimensional model of schizotypy.