The MAFS was developed by clinical consensus, resulting in a 23-item self-report questionnaire with three distinct subscales: general functioning, family-related functioning, and peer-related functioning. MAFS data were collected in a general population sample (N = 842; mean age = 15.0 years [standard deviation = .4]) at baseline and again at 1- and 3-year follow-up. Psychometric analyses included confirmatory factor analysis, calculations of internal consistency, scale correlations, and correlations with the abridged General Health Questionnaire.
Confirmatory factor analysis showed that the hypothesized 3-factor structure fits well to the MAFS data. All scales showed adequate internal consistency (greatest lower bound: .75-.91) and sufficient discriminative ability (scale intercorrelations: ¦Ñ = .15-.52). Of the scales, general functioning was most strongly correlated with the General Health Questionnaire, whereas family- and peer-related functioning showed weaker correlations with this general measure. The results were stable across repeated measurements and gender groups.
The MAFS is an easy-to-use instrument with good psychometric characteristics, which could be suitable for a broad range of future research applications, especially when a multidimensional and unbiased indication of normal adolescent functioning is required.