pequod (version 0.0-5)

lmres: Moderated regression with residual centering

Description

Fit moderated linear regression with both residual centering and mean centering methods.

Usage

lmres(formula, data, residual_centering, centered, ...)
"lmres"(formula, data, residual_centering=FALSE, centered = "none", ...)

Arguments

formula

an object of class "formula": a symbolic description of the model to be fitted.

data

a data frame

centered

variables wich must be centered

residual_centering

"FALSE" generate a moderated using standard lm regression, "TRUE" generate a moderated regression with residuals centering

...

Value

lmres returns an object of class "lmres".An object of class "lmres" is a list containing at least the following components:

regr.order

the numeric order of the fitted linear model

formula.StepI

the formula of the first order regression

formula.StepII

(only where relevant) the formula of the second order regression

formula.Stepfin

the formula of the x (max(x)=3) order regression

beta.StepI

a named vector of standardized coefficients for the first order regression

beta.StepII

(only where relevant) a named vector of standardized coefficients for the second order regression

beta.Stepfin

a named vector of standardized coefficients for the x (max(x)=3) order regression

StepI

a lm object for the first order regression

StepII

(only where relevant) a lm object for the second order regression

Stepfin

a lm object for the x (max(x)=3) order regression

F_change

is a list containing F change statistics

Details

Moderated regression without residual centering : For any interaction term, the product is computed and entered in the final model. In order to perform a mean centered moderated regression, predictors must be centered Moderated regression with residual centering: For any interaction term with order n, a regression with low order terms (n-1) is computed, and Interaction residuals are entered in the final model.

References

Little, T. D., Bovaird, J. A., & Widaman, K. F. (2006). On the Merits of Orthogonalizing Powered and Product Terms: Implications for Modeling Interactions Among Latent Variables. Structural Equation Modeling, 13(4), 497-519.

Cohen, J., Cohen, P.,West, S. G.,&Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.

See Also

“summary.lmres”

Examples

	
	## moderated regression with mean centering
	library(car)
	data(Ginzberg)
	model1<-lmres(adjdep~adjsimp*adjfatal, centered=c("adjsimp", "adjfatal"),
	data=Ginzberg)
	
	## moderated regression with mean centering
	library(car)
	data(Ginzberg)
	model1<-lmres(adjdep~adjsimp*adjfatal, centered=c("adjsimp", "adjfatal"),
	data=Ginzberg)
	## moderated regression with mean centering
	model2<-lmres(adjdep~adjsimp*adjfatal,residual_centering=TRUE,
	centered=c("adjsimp", "adjfatal"), data=Ginzberg)
	## three way interaction with mean centering
	library(car)
	data(Highway1)
	model3<-lmres(rate~len*trks*sigs1, centered=c("len","trks","sigs1"),data=Highway1)

## The function is currently defined as
function (formula, data, centered, ...) 
UseMethod("lmres")