A Comparison between the Use of Beta Weights and Structure Coefficients in Interpreting Regression Results

Autor/in	Tong, Fuhui
Titel	A Comparison between the Use of Beta Weights and Structure Coefficients in Interpreting Regression Results
Quelle	(2006), (18 Seiten) PDF als Volltext kostenfreie Datei Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; online; Monographie
Schlagwörter	Researchers; Predictor Variables; Regression (Statistics); Comparative Analysis + Suchen Sie Ihr Suchwort? Researcher; Forscher; Prädiktor; Regression; Regressionsanalyse
Abstract	Background: An extensive body of researches has favored the use of regression over other parametric analyses that are based on OVA. In case of noteworthy regression results, researchers tend to explore magnitude of beta weights for the respective predictors. Purpose: The purpose of this paper is to examine both beta weights and structure coefficients in interpreting regression results. Data Collection and Analysis: Two heuristic examples will be illustrated. Findings: When predictor variables are non-correlated, their beta weights are equal to the corresponding structure coefficients. However, when predictors are correlated with each other, a variable with zero beta weight had a strong structure coefficient (r = 0.91). Therefore, failure to compute structure coefficients may cause erroneous interpretation in regression, when multicollinearity is present between predictors. Conclusion: Since structure coefficients are not affected by multicollinearity, by consulting both beta weights and structure coefficients, researchers can report unbiased and valid regression results by balancing their attention to both interpretation perspectives. Chapters 8 and 9 in Thompson (2006) provide more detail on these various issues. Citation: Tong, F. (2006, February). "A comparison between the use of beta weights and structure coefficients in interpreting regression results". Paper presented at the annual meeting of Southwest Educational Research Association, Austin, TX. (Contains 3 tables and 2 figures.) (Author).
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2017/4/10