Abstract
The construct validity and generalizability of the Mental Health Statistics Improvement Program (MHSIP) adult consumer satisfaction survey was examined using a representative sample (N = 4242) of public mental health client surveys and assessment data. Factor analysis found five factors generally supporting the hypothesized scales of Satisfaction, Outcomes-Functioning, Access, Quality, and Social Connectedness that had acceptable reliability (average coefficient alpha = 0.91). The relation of the satisfaction survey scales to clinician-rated symptom scales, hospitalization, and functioning for 3 years prior to and after the survey was examined. Correlations were modest in magnitude but were in the predicted directions, with greater patient satisfaction being associated with lower symptoms and more positive outcomes. The largest predictors of the clinician-rated symptom scales were Outcomes-Functioning and Social Connectedness.