Theory-Testing, Generalization, and the Problem of External Validity

Abstract

External validity refers to the generalization of research findings, either from a sample to a larger population or to settings and populations other than those studied. While definitions vary, discussions generally agree that experiments are lower in external validity than other methodological approaches. Further, external validity is widely treated as an issue to be addressed through methodological procedures. When testing theories, all measures are indirect indicators of theoretical constructs, and no methodological procedures taken alone can produce external validity. External validity can be assessed through determining (1) the extent to which empirical measures accurately reflect theoretical constructs, (2) whether the research setting conforms to the scope of the theory under test, (3) our confidence that findings will repeat under identical conditions, (4) whether findings support the theory being tested, and (5) the confirmatory status of the theory under test. In these ways, external validity is foremost a theoretical issue and can only be addressed by an examination of the interplay between theory and methods.