Using the statistical method of Free Choice Profiling to evaluate Health in Dairy Cattle

Summary

The aim of this study was to test the application of a novel statistical method, Free Choice. Profiling (FCP), for evaluation of cattle health. Participants were two groups of veterinarians: a group with biomedical background (n=7) and a group, who practiced homeopathy (n=8). Examined animals were 10 dairy cows. Using FCP methodology participants evaluated behavior in relation to health by watching video sequences of the cows, and evaluated health status by conducting clinical examinations. Data was analyzed with Generalized Procrustes Analysis, both separately for the 2 groups and for the merged group. Significant agreement was found between participants for all analyses (p<0.001), particularly regarding behavior (between 70-80%). Semantic interpretation of the 2 main dimensions of consensus resulted in the behavioral descriptors: ‘robustness/affected by human presence’ and ‘activity’. Descriptors for the clinical examination referred to ‘usage’ characteristics. Good agreement was seen regarding positioning of the animals according to the main dimensions for the video analyses. Correlation between individual somatic cell counts and positions of the cows on the 1st dimensions of the analyses was low (r<0.3), except for merged clinical analysis (r=0.7). Given the ability of these analyses to detect semantically interpretable patterns, it was concluded that FCP might, with further research, provide a useful method for the exploration of health evaluations.