When causal inferences are to be synthesized across multiple studies, efforts to establish the magnitude of a causal effect should be balanced by an effort to evaluate the generalizability of the effect. The evaluation of generalizability depends on two factors that are given little attention in current syntheses: construct validity and external…

Presents a prototypical case illustrating what positivist research means and argues that even critics of positivist research are logically committed to propositions that can be tested only through positivist research paradigms. The relationship between the nature of a community's research enterprise and its educational provisions must be causal.…