New method accurately estimates variance components in linear mixed models.
The spectral decomposition estimate (SDE) is a new method for estimating fixed effects and variance components in linear mixed models. This study focused on SDE in models with two variance components. The researchers found conditions where the analysis of variance estimate equals the SDE of variance components. They also showed that the SDE of variance components is the most accurate estimate under certain circumstances.