High Records Contrasted with Unselected Records and with Average Records as a Basis for Selecting Cows

Abstract

While making some preliminary studies of suitable ways to test the relative values of production records, singly and in various combinations, for predicting future records of the same cow and for evaluating her transmitting ability, we noted an article by Copeland (2) purporting to show some advantages for using a cow's highest record as an index to her lifetime producing ability and to her transmitting ability. As our data, which came from a different source, seemed appropriate for testing Copeland's conclusions, we repeated a part of his analysis on our own material and added a few other features which, it is hoped, will demonstrate why the ex post facto selection of the highest record is an unsound procedure a procedure more likely to lead to fallacious results than if an unselected record or, better still, the average of all a cow's records were used as the measure of her ability.