Abstract
Data from performance tested dual-purpose bulls born between 1969 and 1974 and from 1977 to 1982 were analysed separately for the influence of breed and non-genetic factors on daily gain. Altogether data on 2028 bulls representing 165 sires were analysed in Period I and 3 189 bulls from 186 sires in Period II. The results showed a significant influence of test station, year and season of birth on daily gain. Interactions between these effects were also demonstrated. The adjustment made in the present performance testing routine for the variation in daily gain due to month of birth by using contemporary groups did not work satisfactorily. The effect of breed was significant. Error variance for daily gain increased from Period I to Period II in most of the stations. In a future application of a Mixed Model Procedure it would appear necessary to include in the statistical model for evaluation the effects of test station, year and season of birth and the interactions between these. Pre-adjustment of daily gain for the effect of month of birth is also recommended.