Comparative research involving American [USA] minorities is examined in terms of external validity. The process of insuring external validity is approached through the use of cultural, interpretive, population, ecological and construct validity. These aspects of validity are combined into mechanistic and teleological metamodels. Population, ecological and construct validity provide a basis for a mechanistic or causal metamodel, while a teleological approach usually emphasizes cultural and interpretive validity. Both approaches are combined into a model which includes an intentional account of behavior set within a causal framework.