Reliability and validity of the interview and self-report versions of the BASIS-32.

The psychometric properties of the interview and self-report versions of the BASIS-32 were compared. A total of 120 severely mentally ill adults enrolled in psychosocial rehabilitation were randomly assigned to either a self-report or an interview condition. The BASIS-32 had good internal consistency and test-retest reliability on most subscales; coefficients were higher in the self-report condition. Only the interview version of the psychosis subscale had unacceptable internal consistency. Validity correlations were generally good for the symptom subscales but disappointing for the functional domains. The subscale scores did not discriminate between diagnostic subgroups.