RESUMO
Cognitive biases have been demonstrated to be important in developing and maintaining psychosis. However, self-report measures for everyday clinical practice have been developed only recently. We aimed to study one of these instruments for assessing cognitive biases: the Davos Assessment of Cognitive Biases Scale (DACOBS). In a Spanish sample of 84 patients diagnosed with schizophrenia-spectrum disorders and 152 healthy controls, we examined a) the factor structure using Confirmatory Factor Analysis (CFA) to test the original one-, three- and seven-factor solutions, b) the reliability (Cronbach's alpha), c) the discriminative power (Multivariate Analysis of Covariance - MANCOVA) and d) the relationships of cognitive biases with positive psychotic-like experiences (PPLEs) in healthy individuals and with psychotic symptoms in schizophrenia-spectrum patients. The CFA revealed that the seven-factor solution achieved the best fit. The DACOBS overall scale (Cronbach's alpha = .92) and subscales obtained good internal consistencies. MANCOVA, controlling for age and education, demonstrated that all subscales differentiated between healthy controls and psychotic patients (Wilks' Lambda = 0.87; F7, 226 = 4.70; p < .000; partial eta squared = 0.13). In addition, the DACOBS showed high correlations with PPLEs (controls) and moderate correlations with positive and general symptoms (patients), demonstrating its predictive validity. Concluding, the DACOBS proved to be a psychometrically suitable instrument for assessing cognitive biases in psychosis and adequately differentiated between patients and healthy individuals within the Spanish population. Norm scores are provided.
