Gjm. Rosa et al., Assessing relationships between genetic evaluations using robust regression with an application to Holsteins in Uruguay, ACT AG SC A, 51(1), 2001, pp. 21-34
Globalization of dairy cattle breeding has created a need for international
sire proofs. Some early methods for converting proofs from one population
to another are based on simple linear regression. An alternative robust reg
ression method based on the t-distribution is presented, and maximum likeli
hood and Bayesian techniques for analysis are described, including the situ
ation in which some proofs are missing. Procedures were used to investigate
the relationship between Holstein sire proofs obtained by two Uruguayan ge
netic evaluation programs. The results suggest that conversion equations de
veloped from data including only sires having proofs in both populations ca
n lead to distorted results, relative to estimates obtained using technique
s for incomplete data. There was evidence of non-normality of regression re
siduals, which constitutes an additional source of bias. A robust estimator
may not solve all problems, but can provide simple conversion equations th
at are less sensitive to outlying proofs and to departures from assumptions
.