The Sisters' Riddle And The Importance Of Variance When Guessing Demographic Rates From Kin Counts

Abstract

Goldman (1978) has proposed a formula derived from stable population theory for estimating growth rates from certain counts of younger and older sisters in a population. Computer microsimulation outputs show extraordinarily large statistical errors for estimation when stable population assumptions are modified to allow for random variability such as would be encountered with field work on small populations. A derivation of Goldman's formula on more interpretable lines than that in Goldman (1978) identifies a source of variance special to this particular formula so that the large statistical errors for this example need not count in general against the feasibility of estimating demographic rates from counts of kin. © 1980 Population Association of America.