Courville, T. & Thompson, B. (2001). Use of structure coefficients in published multiple regression articles: β is not enough, Educational and Psychological Measurement, 61(2), 229-248.

Abstract

PURPOSES: To introduce an alternative interpretation of multiple regression analysis. METHODS: A documentary and critical interpretation research. RESULTS: Examination in sample articles demonstrated that there were misinterpretations of results from multiple regression analysis when regression weights alone were taken into consideration for both prediction and explanation or theory testing purposes, particularly, when researchers deem unimportantly any predictor variables with near-zero weights. THEORETICAL/POLICY IMPLICATIONS: In the interpretation of multiple regression analysis, the researchers in both social and natural sciences are advised to consider both the regression weights and structure coefficients to avoid misinterpretation of the prediction and explanation.