A method in human genetics for comparing the proportion in families of children in which a proband shows a specific trait with the proportion expected if the trait were inherited as a single gene. For example, if one considers a group of families, each with both parents heterozygous for a recessive gene and each with two children, the proportion of affected children is 57%, not 25%. This is because the families are chosen in the first place through an affected child, and all sibships in which just by chance no affected individuals occurred have been left out. Thus, there is an ascertainment bias that loads the results in favor of the trait. See Chronology, 1910, Weinberg.
Subjects: Genetics and Genomics.