Alternative modeling of fixed effects in test day models to increase their usefulness for management decisions

Current genetic evaluation models and especially fixed effects are setup in order to minimize bias in genetic evaluations but not for optimal usefulness for management. This is especially true for the definition of contemporary group effects. Very moderate alternative modeling of this effect as the sum of fixed herd-year of test and herd-month of test over a time period and a random herd test day deviation increased their usefulness for management nearly without producing ranking changes in animals (rank correlations > 0.99) for milk, fat and protein yields.