We will follow the motivational example provided by Robinson (1991) Let’s assume we have
collected data on lactation yields of dairy cows in three different herds. Let’s assume that the cows’ sire genetic merit is treated as a random effect while herds are treated as fixed.