Statistical approaches to interpret multiple regression results

Abstract

The importance of taking both β weights and structure coefficients in interpreting regression studies, especially in applied linguistics papers, has often been ignored. The purpose of the present study was to explain both the regression coefficient and the structure coefficient to support the view that the inclusion of these two coefficients will augment the interpretation of results. Some examples of misinterpretations in regression studies were provided. Then the author applied both β weights and structure coefficients in interpreting the results of an experimental study in applied linguistics. β weights and structure coefficients were applied to investigate the predictive power of three predictors on the grammatical acquisition of a pair of twins. The results highlighted the importance of both coefficients in multiple regression studies.