The construct and predictive validity of adaptive behavior of mentally retarded persons was investigated using a multitrait multimethod analysis. The validity issues of concern were (a) how generalizable the adaptive behavior dimension is over a variety of measurement methods, (b) how adaptive behavior relates to mental age (MA), and (c) how well adaptive behavior assessment references placement in programs and situations. Adaptive behavior and MA were assessed by program participation, psychometric ratings, and overall staff ratings. Adaptive behavior had high convergent validity and low discriminant validity and predicted MA program placement better than did MA. We concluded that adaptive behavior scores could replace MA.
