A confirmatory factor analytic evaluation of the pentagonal PANSS model.

The positive and negative syndrome scale (PANSS) is widely used in psychiatric research. Reflecting this common use, considerable attention has been applied to the psychometric properties of this instrument. However, despite the publication of numerous studies and analyses, it remains uncertain how best data from the PANSS should be analysed to best model the symptoms of schizophrenia. A resolution to these concerns seemed to be offered following the publication in 1997 of a large multisite factor analysis that produced the 'pentagonal model', which has subsequently been included in the 2000 revision of the PANSS user manual. However, to date, an independent confirmatory analysis of this model has not yet been published. The aim of this study was to test this model in a new independent sample with confirmatory factor analysis (CFA). Independent confirmation of the fit of the model is required to ensure that its implementation is informed by confirmation of its psychometric properties. CFA was performed in a sample of 347 subjects with schizophrenia. The analysis found that the model had inadequate goodness of fit. The use of the pentagonal model has similar difficulties as earlier models and more research is required to ascertain the optimal method for measuring symptom dimensions in research and clinical settings.