| Abstract: | Data on 2712 bulls tested at station were analysed to study the effect of certain external variables on growth traits of bulls. Most of the animals were Ayrshires (2136). Otherwise Finncattle bulls and Frisian bulls were involved. The testing covers the interval 60–365 days of age, although in part of the data the period is 30–365. Bulls were weighted every 30th day (between 1965 and 1970), or at the ages of 60 and 180 days, and thereafter every 30 days up to the age of 365 days (between 1971 and 1977). Also birth weight was available for 1250 calves. Feeding and general treatment of bulls at station was as uniform as possible for all animals. Conseqúently it can be assumed that effects of non-genetic factors on the growth traits of bulls are minimized. One of the objectives of this study was to compare the growth curves for the three breeds. In our data measurements at certain time-points have dropped out during the course of data collection. Therefore we study the influence of deleting measurements. We also consider the problem of influential measurements at both the design level and the inference level. Several types of growth curves are fitted to the data in order to find the most appropriate model. |
