Subjects were cautioned that the study (1) was not a treatment, (2) was not intended to improve schizophrenia symptoms, (3) carried the possible risk of symptoms worsening, relapse, and hospitalization, and (4) was not a sanction to use alcohol. Subjects were required to pass a questionnaire about the study risks and benefits. Collaterals were involved in the consent process, inclusion was contingent on the clinician's approval, and an ombudsperson (clinician) was appointed to serve subjects.