[R-sig-ME] help with MANOVE with mixed effects

David Duffy David.Duffy at qimr.edu.au
Mon May 21 02:16:59 CEST 2012


On Sat, 19 May 2012, David Costantini wrote:

> Dear All I wonder if anyone can help with this issue? I have two 
> fixed factors (experimental group and sex) and a random factor (brood). 
> I need the random factor to control for pseudoreplication because each 
> brood provides a certain number of siblings, which are 
> pseudoreplicates because share genes/environment. Then I have 8 
> variables. so I would like to to compare the covariance structures among 
> groups while controlling for the random factor. I was told that R might 
> do that, I mean a manova with mixed effects (manova(lme(....))

Not as easily as that ;) One can represent a multivariate model as a lmer 
or lme univariate model with appropriate random effects representing each 
item (there is recent query about this on the list).

MCMCglmm will fit as a multivariate model.

Or you can use structural equation modelling packages eg OpenMx.

Or you can go outside of R to other packages eg WOMBAT 
(http://didgeridoo.une.edu.au/km/wombat.php), MENDEL, ASReml...

-- 
| David Duffy (MBBS PhD)                                         ,-_|\
| email: davidD at qimr.edu.au  ph: INT+61+7+3362-0217 fax: -0101  /     *
| Epidemiology Unit, Queensland Institute of Medical Research   \_,-._/
| 300 Herston Rd, Brisbane, Queensland 4029, Australia  GPG 4D0B994A v



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