Heightened concerns over the 1992 AAMR definition: advocacy versus precision.

We reiterated our earlier concerns (MacMillan et al., 1993) regarding the 1992 AAMR definition of mental retardation and replied to a defense of the definition (Reiss, 1994). Our major point here is that definitional precision should not be sacrificed to advance a particular ideological position. The result of such a compromise is a definition lacking precision and necessitating identification dependent upon assessment of behavioral dimensions for which there is currently no reliable measurement. We explored the varied purposes served by classification systems and criticized the new AAMR definition for lack of precision needed to serve research. In addition, we attempted to reduce confusion over the concepts of reliability, assessment, test bias, and use of factor analysis to establish the independence of adaptive skill domains. Some factual errors and incorrect representations in Reiss's article were pointed out.