BackgroundEvidence indicates that the neuropathology of auditory verbal hallucinations (AVH) varies according to their phenomenological characteristics. Therefore, AVH should be subgrouped accordingly in hallucinations research. As evaluation of these characteristics depends entirely on the patient report, obtaining measurement of the reliability of these reports is crucial.MethodA computerized binary scale of auditory speech hallucinations (cbSASH) was developed to evaluate the phenomenology of AVH. It includes two subscales (inconsistency and malingering) to assess the reliability of the patient report. The cbSASH was administered along with MMPI-2, a general psychopathology scale, which includes similar validity subscales. Thirty-four psychotic patients with history of AVH were enrolled in this study.ResultsThe scores on the inconsistency and malingering subscales of the cbSASH were correlated with the scores on the corresponding validity subscales in the MMPI-2. The combination of the malingering and inconsistency subscales provided robust measures of the reliability and ability of the patients' descriptions of their hallucinations.ConclusionThe cbSASH provides a reliable and comprehensive evaluation of the phenomenology of AVH. Consequently, it is possible to subgroup patients according to the characteristics of their hallucinations. This refinement of AVH phenotypes could reduce the noise and inconsistency noted in AVH and psychosis research.