Application of Regression Analysis Concepts to Retrospective Research in Speech Pathology

Abstract

The underlying logic of retrospective research in speech pathology is examined critically. This research model focuses on differences among discrete groups. Regression analysis concepts are suggested as a tool to evaluate the practical or theoretical importance of statistically significant differences obtained in retrospective research. This type of analysis focuses on the degree to which a criterion measure can be predicted from knowledge of group classification, rather than on differences among means.