The dominant approach to research in learning disabilities utilizes experimental and quasi-experimental designs to identify the most effective instructional strategies for students with learning disabilities. Research is always undertaken from a point of view, yet the discourse on "what works?" is generally silent on how theoretical frameworks inform all research, including intervention research in learning disabilities. This paper argues that the quality of research in learning disabilities depends on the quality of the theoretical frameworks that inform any research study, including the questions the researcher asks, the methodology employed, and the way the data are analyzed. An example of classroom language research is used to illustrate how a study informed by a sociocultural model of learning demands an alternative to the experimental and quasi-experimental methodologies that dominate research in learning disabilities. (Contains 2 notes.)