Bennett Dyke

Assortative mate choice and mating opportunity on Sanday, Orkney Islands

Abstract A method is presented for the analysis of deviations from random mating. Kinship, demographic, social, and spatial characteristics observed among married couples have been compared with the distributions expected if mates were chosen at random from all possible pairs of mates. This procedure has been used to investigate both failure to mate and patterns of assortative mating for cohorts born on Sanday, Orkney Islands, between 1885 and 1924. Differences in mating opportunity were observed. The 315 males who eventually married on the island differed from the 446 never‐married males in birth order and sibship size as well as geographic and kinship “distances” measured between ego and all females available for marriage. Comparison of wives with the potential mates of married males indicated that mating was assortative with respect to kinship, demographic, social, and geographic characteristics. Further implications of this nonrandom pattern of mate choice are discussed and application of this method ...

ESTIMATION OF CHANGING RATES BY SIMULATION

In a recent publication we have reported the use of computer simulation as a means of estimating vital rates in a small population isolate (Dyke and MacCluer 1973). In that paper we estimated mortality rates using both deterministic and stochastic simulation procedures. Then, incorporating these mortality rates in a more complex stochastic model developed by MacCluer (this volume), we made estimates of age-specific fertility rates. At that time much of our effort was spent in validating the model which had previously been used with non-European and model populations (MacCluer 1967, 1973; MacCluer and Schull 1970a, 1970b; MacCluer et dl. 1971), and in matching selected demographic characteristics of the simulated population with those of the isolate. Much of this procedure was carried out under the assumption of constant vital rates applied to the artificial population over the entire 50-year period simulated. However, it was shown that a better match of final population structures could be achieved when simulation input fertility rates were changed periodically through time. This paper reports the details of the latter procedure, that is, changing simulation input fertility rates through time, and gives estimates of secular change in fertility which result.