Repeatability of Threshold Traits

Publisher Website
Google Scholar
Abstract
Several easily calculated estimators of repeatability of a threshold trait are compared. An estimator due to Pearson (rediscovered by Plackett) of the correlation between scores on the underlying scale is shown to perform well over a number of threshold locations and true correlation values. An inverse prediction equation is calculated to improve it. An estimator due to Lush is a poor estimator of correlation on the underlying scale, but it is noted that this is not its intended use. Rather it is correct for predicting most probable producing ability. Many traits of interest to quantitative geneticists are restricted to only 2 phenotypic classes even though they are expressions of the genetic effects of many loci and may be influenced by several sources of environmental variation.

This publication has 2 references indexed in Scilit: