finding outliers in mixed model

Question

I'm trying to find outliers in this mixed model:

m1 <- lmer(y ~ service + lectage + studage + (1|d) + (1|s), data=InstEval)

So I used the merTools REsim function

reEx <- REsim(m1)

and plotted reEx

p1 <- plotREsim(reEx)

but how can I find which observations are the actual outliers?

score 1 · Answer 1 · answered Jan 16 '19 at 08:51

This is a contentious topic and likely to attract varied answers. Anyway, you could use Cook's distance as in regular OLS as a measure of influence

library(lme4)
mod=lmer(mpg~cyl+disp+hp+(1|drat),data=mtcars)
library(influence.ME)
inf=influence(mod,obs=T)
plot(inf,which="cook")

1 Answers1