Abstract
The expected rate of improvement from selection generally has been based on (1) the intensity of selection applied each generation, (2) the degree of heritability and (3) the interval of time between generations, considering each of the desired traits independently. As actually applied in recent swine breeding experiments, selection has been based on some sort of a weighted average of the more important characters, and has been accompanied by mild inbreeding. The effectiveness of this selection can be evaluated from (1) the average time trend for each character within experimental strains, adjusted for the decline due to increased inbreeding, (2) comparison of strain crosses with representative purebreds and from (3) comparison of topcrosses with representative purebreds. Time trends in performance are misleading as measures of genetic change to the extent that they are influenced by time trends in nutrition, disease, management, climate, or other factors. Nevertheless, they provide a measure of improvement achieved. Copyright © . .