Displaying 1 result from an estimated 1 matches for "speciesmeans".
2009 Jul 16
0
how to get means and confidence limits after glmmPQL or lmer
...d CIs for the levels of treatment on the original
scale. Below I describe some of the things I have tried.
nn<-intervals(mm)#to get confidence intervals on estimates
oo<-nn$fixed
oo[2,]<-oo[1,]+oo[2,]
oo[3,]<- oo[1,]+oo[3,]
pp<-exp(oo)
But the means are not the right means (see speciesmeans below), the
intercept does not correspond with level 1 of treatment (b1), and the
confidence limits seem too big. Also, I'm not sure if I've
backtransformed correctly.
I've also tried directly calculating the means and CIs, but I'm not
sure if the first line below is actually...