[R-sig-ME] Relative importance values generated by MuMIn

[Luca Borger]

Hello,

according to the MuMIn help page for 'model.avg' this is correct:

"With method = "0" (default) all predictors are averaged as if they were 
present in all models in
the set, and the value of parameter estimate is taken to be 0 if it is 
not present in a particular model.
If method = "NA", the predictors are averaged only over the models in 
which they appear."

HTH

Cheers,

Luca


-------------------------------------------
Luca Borger
Postdoctoral Research Fellow
CNRS - Centre d'Etudes Biologiques de Chizé
Villiers-en-Bois
79360 Beauvoir-sur-Niort
France

Tel: +33 (0)549 09 96 13
Fax: +33 (0)549 09 65 26

email: lborger at cebc.cnrs.fr
http://www.researcherid.com/rid/C-6003-2008
http://cnrs.academia.edu/LucaBorger
-------------------------------------------


Le 04/08/2011 13:04, Thackeray, Stephen J. a écrit :
> Dear all,
>
> I am using the MuMIn package for multi-model inference (based upon sets on lmer models) for the first time and have a question about the interpretation of the relative importance values given for variables when using the model.avg function. If you use get.models to select only models with delta AIC<=4 or a 95% confidence set then it is likely that, in this model set, each predictor will appear in a different number of individual models. Do the calculations of relative importance allow for this so that the support for different predictors can be meaningfully compared, even if each appears in a different number of models? In the model.avg call I specify method="0" so that when predictors are absent from any single model they are assumed to actually be present with a parameter (slope) estimate of zero e.g.
>
> TCB.mod.av.95CS<-model.avg(TCB.models2,method="0",rank="AIC",alpha=0.05)
>
> I presume that this means that each predictor is treated as being present in all models in the top model set, so that the relative importance values can be compared across the predictors meaningfully. Do I understand this correctly?
>
> All the best
>
> Steve
>
> Dr Stephen  Thackeray
> Lake Ecosystem Group
> Centre for Ecology and Hydrology
> Lancaster Environment Centre
> Library Avenue
> Bailrigg
> Lancaster
> LA1 4AP
> sjtr at ceh.ac.uk<mailto:sjtr at ceh.ac.uk>
>
>
>



__________ Information from ESET Mail Security, version of virus signature database 6349 (20110804) __________

The message was checked by ESET Mail Security.
http://www.eset.com