Repeated pairs (RP) of surnames of married couples reflect the genetic structure of the population. The number of pairs of husband and wife that match the same two surnames in another couple serves to estimate the inbreeding through size and subdivision of the population. RP = S[Sij(Sij-l)]/N(N—1), in which Sy- is the number of couples with the ith and jth surnames respectively and N is the sum of Stover all surname pairs. The frequency of such repeated instances is then compared with frequencies from repeated runs after randomizing the order of wives and arbitrarily pairing their names with those of husbands. In a census of Paracho, Mexico, among 757 married couples there were 2865 pairs of surnames with a mean rate of repetitions of .000176 which is 25% more than random expectation. In a comparison limited to the 365 Paracho-endogamous couples, the mean rate of repetitions rises to .000303 but the random rate rises even more to .000332, and the excess of observed over random disappears. The numbers of instances of repeated marriages are greater than those of marriages between persons of the same surname, so the standard errors of RP tend to be smaller than those of marital isonymy.
Lasker, Gabriel W. and Kaplan, Bernice A.
"Surnames and Genetic Structure: Repetition of the same Pairs of Names of Married Couples, a Measure of Subdivision of the Population,"
3, Article 12.
Available at: https://digitalcommons.wayne.edu/humbiol/vol57/iss3/12