I swear there was a point in writing an introduction to covariance structures: now we can start joining all sort of analyses using very similar notation. In a previous post I described simple (even simplistic) models for a single response … Continue reading →