The formulation of the linear mixed model is very general and includes as special cases models where only fixed effects or only random effects are used. When only fixed effects are used, this model corresponds to a fixed effects analysis of variance (ANOVA) where parameters can be estimated with ordinary least squares. When random effects are specified, the variance terms can be estimated with maximum likelihood or restricted maximum likelihood (REML) [37]. Since REML does not directly estimate parameters for fixed effects, these coefficients are estimated after the fact by plugging in estimates for the variance components [29].
