Mean-Centering Does Not Alleviate Collinearity Problems in Moderated Multiple Regression Models

By Raj Echambadi, James D. Hess. 2007

485 citations recorded by the source corpus.

Subfields: Food Science, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/843fd7d6af80cbb6d8126f4bb8109ad30128bbfe

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Food Science

In Statistics and Probability

Citations in the corpus

A Comparison of Approaches for the Analysis of Interaction Effects Between Latent Variables Using Partial Least Squares Path Modeling (2010, background): “Although such a proceeding does not necessarily diminish the multicollinearity resulting from building the product term (see Echambadi & Hess, 2007 ; contrary to Cohen, 1978 ; Cronbach, 1987 ), it does facilitate the interpretation of the interaction model results.”
Entrepreneurial Orientation, Firm Performance, and the Moderating Role of Transformational Leadership Behaviors (2015, background): “Echambadi and Hess (2007) recommend re-estimating the regression models for 10 iterations, in each case with a randomly drawn subsample of 90% of all data sets in the pooled sample, arguing that multicollinearity will result in unstable regression coefficients.”
Hidden Badge of Honor: How Contextual Distinctiveness Affects Category Promotion among Certified B Corporations (2017, methodology): “We mean centered all four independent variables, so as to facilitate interpretation of our interaction terms (Echambadi & Hess, 2007).”
Impacts of Hedonic and Utilitarian User Motives on the Innovativeness of User-Developed Solutions: Motives for User-Developed Solutions (2015, unknown): “For ease of interpretation, the constituent variables have been mean-centered (Cohen, Cohen, West, and Aiken, 2003; Echambadi and Hess, 2007).”
How does product program innovativeness affect customer satisfaction? A comparison of goods and services (2011, methodology): “To overcome potential problems due to multicollinearity (Aiken and West 1991), we ran stability tests ( Echambadi and Hess 2007; Homburg et al. 2007a).”
PLease do not answer if you are reading this: respondent attention in online panels (2018, methodology): “We assessed significance using bootstrapping to accommodate for the presence of an interaction term (Echambadi and Hess 2007) 18 Mark Lett (2018) 29:13–21”
Cautions Regarding the Interpretation of Regression Coefficients and Hypothesis Tests in Linear Models with Interactions (2012, background): “The need to center (or standardize) X and M in a regression model including XM as a predictor is a myth that doggedly persists in spite of having been repeatedly debunked ( Cronbach, 1987 ; Echambi & Hess, 2007 ; Friedrich, 1982 ; Hayes, 2005 , pp.”
Lateral Collinearity and Misleading Results in Variance-Based SEM: An Illustration and Recommendations (2012, unknown): “(1) 21 1 i i R VIF − = Situations in which variables are correlated yet not collinear are actually quite common in multivariate analyses (Echambadi & Hess, 2007; Hair et al., 2009).”
The Impact of Corruption on Firm Tax Compliance in Transition Economies: Whom Do You Trust? (2013, background): “8 Interaction terms were not mean-centered, since research suggests that mean-centering of variables does not reduce multicollinearity (Echambadi and Hess, 2007).”
Some Common Myths About Centering Predictor Variables in Moderated Multiple Regression and Polynomial Regression (2012, result): “For example, recent work by Echambadi and Hess (2007) demonstrates some of the results outlined here.”
When does investment in political ties improve firm performance? The contingent effect of innovation activities (2014, methodology): “We used mean-centered variables to ease the interpretation of main and interaction effects in moderated regression models (Echambadi & Hess, 2007; Irwin & McClelland, 2001).”