Answers are grounded in this paper and cite their sources.
A cow index including data on the cow, her sire, dam, and maternal grandsire was correlated more highly (8%) with son's progeny test in Holsteins than was an index from only the cow and her sire. The four other breeds also showed increases with added data. The four-source index showed little or no advantage over the two-source index in predicting daughter performance. Part of that result may be attributed to an environmental correlation between cows and their daughters. Cow data were more important than expected and sire data less important than expected in predicting daughter performance. The reverse was true in predicting son's progeny test. Standard deviations of indexes with four sources of information were 7.6 to 11.6% higher than those of indexes from only the cow and her sire. Correlations between the two-source index and change in index from two to four sources were .05 to .17. A direct approach to computing weights for use in the four-source index required only 5% of the computing time of an approach that required inversion of a matrix of order three. Both approaches
No takes yet. Share an insight, caveat, or question.
Powell et al. (1981) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: