[R-sig-ME] QUESTIONS ABOUT THE PARAMETERS AND EFFECT SIZE OF THR RANDOM EFFECT IN MIXED LINEAR MODEL

Rolf Turner r@turner @end|ng |rom @uck|@nd@@c@nz
Sun Oct 10 22:36:32 CEST 2021


On Sun, 10 Oct 2021 19:01:36 +0800
董龙伟 <donglw17 using lzu.edu.cn> wrote:

> Dear Rolf Turner and other friends
> 
> Thank you for friendly help, and sorry for the vague expression.
> I build the mixed linear model using the lme4, for example
> 
> f1 =lmer(y~x1+x2+x3+(1|r1)+(1|r2),data = dat)
> Note: x1, x2 and x3 are fixed factors, r1 and r2 is the random
> factors. the code and the data been loaded as an attachment
> 
> My questions is how to calculate the parameters and effect size of
> the random effect (r1 and r2). Best wishes to you.

Random effects are not parameters; they are (as the name would imply)
random variables.   The ranef()  function from lme4 will extract the
modes of the instances of these random variables, associated with your
model.  The function VarCorr() will provide estimates of the standard
deviations of these random variables.

I know next to nothing about effect sizes, so I do not have the
expertise to explore your question any further.  Others on
the r-sig-mixed-models list will have such expertise, but unfortunately
the attachment that you sent will not have gone through to the list.
The list mail-handler strips most attachments from email (as a security
measure.

Moreover the "code and result.txt" file in the zip archive that I
received was empty --- which is not much use.

Instead of a zip archive you should send your data and code as separate
text files (with a .txt extension) --- these should go through.  Or you
could simply include your code in the body of the email, and include
the data as well (as long as the data set is not too huge), using
dput().

cheers,

Rolf Turner

-- 
Honorary Research Fellow
Department of Statistics
University of Auckland
Phone: +64-9-373-7599 ext. 88276



More information about the R-sig-mixed-models mailing list