[R-sig-ME] help random effects unested
HDoran at air.org
Mon Jun 30 13:38:48 CEST 2008
Like Andrew, I too am unclear. But, to me it sounds as though you want to treat the effects as uncorrelated and not as crossed per se, but I could be wrong.
First, move to lmre and not lme. There is better support for the more recent package. Second, if you want uncorrelated random effects, please see the linear mixed effects article in R News which shows exactly how to do this.
> -----Original Message-----
> From: r-sig-mixed-models-bounces at r-project.org
> [mailto:r-sig-mixed-models-bounces at r-project.org] On Behalf
> Of Andrew Robinson
> Sent: Friday, June 27, 2008 10:18 PM
> To: Maria Eva Gongora
> Cc: r-sig-mixed-models at r-project.org
> Subject: Re: [R-sig-ME] help random effects unested
> Hi Maria,
> it sounds like you would like those factors to be crossed.
> It is possible to cross factors within lme but it is not easy
> - you might find some guidelines searching R-help via Google.
> Alternatively you could try to use the lmer() function
> within the lme4 package, this function allows you to cross
> random effects.
> I hope that this helps,
> On Sat, June 28, 2008 2:13 am, Maria Eva Gongora wrote:
> > I am using mixed effect models (library nlme, lme) to evaluate
> > factors that influence bycatch of hake in a shrimp fishery.
> Data are
> > collected though an on-board observer program and I
> decided to treat
> > the vessel id and the observer id as random effects. A
> priori I would
> > expect the observer effect to be independent of vessel.
> However, I
> > did not find a way to incorporate both factors as independent,
> > unnested random effects; I could only estimate the effects
> as "vessel" and "observer in vessel"
> > (nested).
> > Is it possible to incorporate two random effects unested?
> > Lic. María Eva Góngora
> > _________________________________________________________________
> > Descargá ya gratis y viví la experiencia Windows Live.
> > [[alternative HTML version deleted]]
> > _______________________________________________
> > R-sig-mixed-models at r-project.org mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
> Andrew Robinson
> Senior Lecturer in Statistics Tel:
> Department of Mathematics and Statistics Fax:
> +61-3-8344 4599
> University of Melbourne, VIC 3010 Australia
> Email: a.robinson at ms.unimelb.edu.au Website:
> R-sig-mixed-models at r-project.org mailing list
More information about the R-sig-mixed-models