jump to main area
:::
A- A A+

Seminars

A Multivariate-t Threshold Model for Clinical Mastitis in Norwegian Cattle

  • 2002-02-22 (Fri.), 10:30 AM
  • Recreation Hall, 2F, Institute of Statistical Science
  • 張幼玫 女士
  • Ph. D. Candidate in Department of Animal and Dairy Sciences, University of Wisconsin–Madison

Abstract

Clinical mastitis records from 30 days pre-calving to 300 days post-partum on 36,178 first-lactation Norwegian Cattle cows were analyzed. Heritabilities and genetic correlations between clinical mastitis in 4 periods of lactation were inferred using a multivariate-t threshold model. This was compared with a multivariate probit model via a pseudo Bayes factor and an analysis of residuals. MCMC algorithms were used to draw from posterior distributions. The pseudo Bayes factor favored the multivariate- t threshold model. This model also had smaller residuals when cows had mastitis. The posterior mean of the degrees of freedom for the t model was 6.4, pointing away from the assumption of a normal liability. In the t model, heritability of liability to clinical mastitis was 0.15 before calving, and ranged between 0.06 and 0.07 after calving. Genetic correlations ranged between 0.52 and 0.75.

Update:
scroll to top