The psychometric properties of the Peters et al. Delusions Inventory (PDI) in Taiwan: reliability, validity, and utility

The Peters et al. Delusions Inventory (PDI) is a commonly used instrument to measure delusion proneness in the general population and includes dimensions that measure distress, preoccupation, and conviction of unusual beliefs. This self-report scale has already been translated into several languages. However, there has not been a validated Taiwanese version previously reported. The aims of the present study were to translate and test the cross-cultural reliability and validity of the PDI in Taiwanese as well as to establish its sensitivity, specificity, and discriminative validity. We administered the questionnaire to a consecutive sample of 253 participants with (n = 154; clinical group including schizophrenia and affective psychosis) or without psychotic disorders (n = 99; non-clinical group). In addition to the Taiwanese version of the PDI (PDI-T), the Taiwanese version of the Brief Psychiatric Symptom Rating Scale (BSRS) was used to measure the severity of psychopathology. We tested the psychometric properties of the PDI-T, including its construct validity, internal consistency, test–retest reliability, concurrent, and discriminative validity. Overall, the PDI-T showed good construct validity, internal consistency, and stability over time, and it was significantly correlated with the BSRS subscales of psychotic symptoms. The convergent and discriminative validity was satisfactory. The area under the receiver operating characteristic curve of the PDI-T was 0.752. This research found that the most appropriate PDI-T yes/no cut-off scores for determining the absence and presence of delusion proneness were 5 and 13. The PDI is a reliable and valid instrument for measuring the dimensionality of delusion proneness and appears to complement subclinical psychosis assessment scales for both epidemiological and clinical research in Taiwan.