Displaying 1 result from an estimated 1 matches for "compfinal".
2012 May 31
1
anova of lme objects (model1, model2) gives different results depending on order of models
...heiro and Bates book or in ?anova, or in Google for that matter
which unfortunately only returns results about ANOVA which isn't much
help. I'm using the latest version of R and nlme, just checked both.
Here is the code and output:
> PHQmodel1=lme(PHQ~Age+Gender+Date*Treatment, data=compfinal,
random=~1|Case, na.action=na.omit)
>
> PHQmodel2=lme(PHQ~Age+Gender+Date*Treatment, data=compfinal,
random=~1|Case, na.action=na.omit,
+ correlation=corAR1(form=~Date|Case))
> anova(PHQmodel1, PHQmodel2) # accept model 2
Model df AIC BIC...