Biomathematics & Statistics Scotland
Lameness in dairy cattle causes distress to the animals and reduces their productivity. Intervention studies offer an opportunity to objectively assess strategies to reduce the incidence of lameness. However, the analysis of data from such studies is complicated by two issues. Firstly, the temporal and within-animal correlations in data collected from different claws of many different animals. Secondly, the effects of variables such as time of year, time since calving and age of the animals are unavoidably confounded. As well as developing a range of analyses using both parametric and additive models, we have established a simulation approach to evaluating the statistical power and cost of different sampling strategies to improve the efficiency of future multi-factorial studies.
Further details from: Sarah Brocklehurst
Article date 2007
