Heritability and genetic correlation: Standard errors of estimates and sample size

Standard errors of heritability as estimated by four different methodologies (regression of offspring on midparent values, regression of offspring on singleparent values, intraclass correlation of full sibs, and intraclass correlation of half sibs) are tabulated. Standard errors of the genetic correlation (correlation between additive genetic values for two characters on the same individuals in a population) as estimated by analogous methods are also presented. These tables suggest that quantitative genetic analyses, including studies of the genetic correlation among characters, should not be undertaken unless resources are available which would allow the testing of at least 400 families of four members each.