It is traditionally assumed that the random effects in mixed models follow a multivariate normal distribution, making likelihood-based inferences more feasible theoretically and computationally.

This is a preview. Log in through your library . Abstract Linear mixed-effects models are a powerful tool for modelling longitudinal data and are widely used in practice. For a given set of covariates ...