Factors affecting accuracy in the evaluation of progeny tests of growth rate in cattle

Two progeny tests for beef, one using 452 calves brought centrally to a test centre, the other using 768 calves commercially fattened on 81 farms, were compared to isolate environmental sources of variation. It was concluded that appreciable reduction of error variance ensued from the use of a growth curve, the maintenance of pen identity, and the removal of variation due to pens and months of birth. The fixed effects model underlying the experimental design on the centre yielded relatively robust estimates of variance components.