[R-sig-ME] Could we calculate the variance explained by each fixed variable when we use linear mixed model?

[Jake Westfall]

The discussion on the wiki is a little different, since it is about getting some global R^2 statistic for the whole model, whereas the OP here appears to want a separate variance explained statistic for each predictor, analogous to partial eta-squared in multiple regression.

Jake

________________________________________
From: R-sig-mixed-models <r-sig-mixed-models-bounces at r-project.org> on behalf of Dan McCloy <drmccloy at uw.edu>
Sent: Friday, December 18, 2015 12:33 PM
To: 黄茹
Cc: R-SIG-Mixed-Models at r-project.org
Subject: Re: [R-sig-ME] Could we calculate the variance explained by each fixed variable when we use linear mixed model?

There is a discussion of variance explained on the Wiki:
http://glmm.wikidot.com/faq
See the section "How do I compute a coefficient of determination..."


On Sun, Dec 13, 2015 at 8:20 PM, 黄茹 <nefuitphuangru at gmail.com> wrote:
>      Can we calaulate the respective variance explained by each fixed
> variable when we use linear mixed model?
>
> Namely, for R code of linear mixed model: model=lmer(y~x1+x2+(1/x3),data),
> x1,x2 are fixed variables, and x3 is random effect. could we compute the
> variance of y explained by x1,x2, respective?
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

_______________________________________________
R-sig-mixed-models at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models