Multiple classification analysis and its application to the 1974 Fiji Fertility Survey

Type Journal Article - World Fertility Survey: Occasional Papers
Title Multiple classification analysis and its application to the 1974 Fiji Fertility Survey
Issue 22
Publication (Day/Month/Year) 1980
Page numbers 111-147
The basic formulation and properties of multiple classification analysis (MCA) are presented along with the relationship of MCA to other related statistical techniques for the study of fertility and demographic data. Such a mathematical treatment was given the Fiji Fertility Survey, 1974, and findings are presented. Fiji's crude birth rate dropped from 40/1000 to 28/1000 in the period from 1950-1973. In a cumulative cohort, fertility, age at marriage, and contraception use were studied by considering 3 types of variables: dependent (number of children ever born), intermediate (age at marriage, marital stability, and use of contraception), and background (residence in childhood, religion, ethnicity, education and labor force participation). It was discovered, among other things, that the higher marital fertility of Indians was significantly attributable to a divergence in nuptiality patterns. 74% of Indian women had used at least 1 method of contraception at some time whereas 59% of Fijians had. In addition to rising age at marriage and wide use of contraceptives, another factor contributing to the downward fertility trend was marital instability (Fijians have higher marital dissolution rates). Among Fijians, there was no consistent difference in fertility observed with respect to women's educational levels or religion; residential classification had a somewhat minor effect on fertility variations. The remainder of the chapter presents statistical results.

Related studies