British Peerage Reanalysis

The paper compares linear regression, two-variable regression, and the Poisson regression used in the Caerphilly study. The main analytic sample included 2,919 women born before 1906 who had exact birth and death dates and survived at least to age 55. The central variables in the reanalysis were age at death, number of children, and year of birth. The peerage sample is limited because it represents British aristocratic family records rather than a general human population. The paper reanalyzes the British peerage genealogical database used in the Westendorp and Kirkwood study.