Context: Generalization is a critical concept in all research designed to generate knowledge that applies to all elements of a unit (population) while studying only a subset of these elements (sample). Commonly applied criteria for generalizing focus on experimental design or representativeness of samples of the population of units. The criteria tend to neglect population diversity and targeted uses of knowledge generated from the generalization. Objectives: This article has two connected purposes: (a) to articulate the structure and discuss limitations of different forms of generalizations across the spectrum of quantitative and qualitative research and (b) to argue for considering population heterogeneity and future uses of knowledge claims when judging the appropriateness of generalizations. Research Design: In the first part of the paper, we present two forms of generalization that rely on statistical analysis of between-group variation: analytic and probabilistic generalization. We then describe a third form of generalization: essentialist generalization. Essentialist generalization moves from the particular to the general in small sample studies. We discuss limitations of each kind of generalization. In the second part of the paper, we propose two additional criteria when evaluating the validity of evidence based on generalizations from education research: population heterogeneity and future use of knowledge claims. Conclusions/Recommendations: The proposed criticisms of research generalizations have implications on how research is conducted and research findings are summarized. The main limitation in analytic generalization is that it does not provide evidence of a causal link for subgroups or individuals. In addition to making explicit the uses that the knowledge claims may be targeting, there is a need for some changes in how research is conducted. This includes a need for demonstrating the mechanisms of causality; descriptions of intervention outcomes as positive, negative, or neutral; and latent class analysis accompanied with discriminant analysis. The main criticism of probabilistic generalization is that it may not apply to subgroups and may have limited value for guiding policy and practice. This highlights a need for defining grouping variables by intended uses of knowledge claims. With respect to essentialist generalization, there are currently too few qualitative studies attempting to identify invariants that hold across the range of relevant situations. There is a need to study the ways in which a kind of phenomenon is produced, which would allow researchers to understand the various ways in which a phenomenon manifests itself.