2016
DOI: 10.3758/s13428-016-0827-9
|View full text |Cite
|
Sign up to set email alerts
|

Mean centering, multicollinearity, and moderators in multiple regression: The reconciliation redux

Abstract: In this article, we attempt to clarify our statements regarding the effects of mean centering. In a multiple regression with predictors A, B, and A × B (where A × B serves as an interaction term), mean centering A and B prior to computing the product term can clarify the regression coefficients (which is good) and the overall model fit R will remain undisturbed (which is also good).

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
56
0
1

Year Published

2017
2017
2024
2024

Publication Types

Select...
10

Relationship

0
10

Authors

Journals

citations
Cited by 102 publications
(57 citation statements)
references
References 7 publications
0
56
0
1
Order By: Relevance
“…Although some significant inter‐correlations between the independent variables were observed, none of the correlation coefficients was above the level considered to be serious, which is generally accepted as 0.80 or higher (Hair et al, ). Iacobucci, Schneider, Popovich, and Bakamitsos () suggest that “mean centering clarifies regression coefficients (that is good) without altering the overall R 2 (that is also good) and helps researchers alleviate both micro and macro multicollinearity” (p. 403). Thus, before the analysis, all the variables were mean centered to minimize the threat of multicollinearity.…”
Section: Methodsmentioning
confidence: 99%
“…Although some significant inter‐correlations between the independent variables were observed, none of the correlation coefficients was above the level considered to be serious, which is generally accepted as 0.80 or higher (Hair et al, ). Iacobucci, Schneider, Popovich, and Bakamitsos () suggest that “mean centering clarifies regression coefficients (that is good) without altering the overall R 2 (that is also good) and helps researchers alleviate both micro and macro multicollinearity” (p. 403). Thus, before the analysis, all the variables were mean centered to minimize the threat of multicollinearity.…”
Section: Methodsmentioning
confidence: 99%
“…Candidate models were tested to residual’s homoscedasticity (Ou et al, 2015). Also, multicollinearity between predictive variables was evaluated (Iacobucci et al, 2017). Furthermore, a correction for optimistic prediction and overfitting was performed according to Harrell’s bootstrapping algorithm (Harrell et al, 1996), which is based on using bootstrapped datasets to internally validate the multiple linear regression models as well as to repeatedly quantify the degree of overfitting in the model-building process.…”
Section: Methodsmentioning
confidence: 99%
“…The final step consisted of the Perceived Entitlement Behavior × LMX interaction term. Cohen, Cohen, West, and Aiken (2013) recommended centering all predictors whereas others contend that centering eases interpretation of results in moderated regression analyses (Iacobucci, Schneider, Popovich, & Bakamitsos, 2017). For these reasons, all predictors were centered.…”
Section: Methodsmentioning
confidence: 99%