Questions to be addressedQuestion 2 As expected, there was a positive association between average milk yield and average concentrate fed during the trial (Figure 3 of Romney et al.). But this is an overall effect and doesn't take into account differences in patterns of feeding at different phases of lactation. It might be worth examining the nature of the association at different stages of lactation to see whether any further light can be thrown on the influence of level of concentrate feeding on milk yield. We shall divide the lactation period into five 7-week phases. The data from both volunteer and control farmers were pooled when it was realised that some control farmers were following the guidelines given to the volunteers. Nevertheless, there was considerable variation in the ways that farmers (both volunteer and control) followed the recommendations for concentrate feeding. It might be possible, therefore, to classify the data set into different levels of concentrate feeding in early and late lactation and to determine average lactation curves for different categories of farmers (low-low, high-low and high-high, say). Therefore the second question we shall address is:
|
