9.4 Summary

This chapter demonstrates, through simulations, how Type I error can get inflated due to (a) an overly simplified random effects structure in linear mixed models, (b) mis-specification of the likelihood, and (c) multiple comparisons.