I am using the rptR package to assess the repeatability of personality tests for individuals tested on different days. The only thing included in the model at present is whether or not the individual emerged during that test (emerged=1 or 0) and the ID of the individual as a random effect.

rptrBinary(emerged ~ (1|ID), grname=c("ID"), data=data, link="logit", nboot=1000, npermut=1000)

When I run the model above, I get an output like this:

Link-scale approximation:

R=0.549

SE=0.323

CI=[0,0.988]

P = 0.00508 [LRT]

= 0.005 [Permutation]

Original scale approximation:

R=0.413

SE=4.228

CI=[0,0.734]

P = 0.00508 [LRT]

= 0.007 [Permutation]

Because the p-values from the likelihood and permutation tests appear significant but the confidence intervals appear very wide, I am not sure how to best interpret these results. If anyone has any insight or resources beyond the original vignette from the authors, that would be appreciated.