Trussell J, Hammerslough C
Demography. 1983 Feb;20(1):1-26.
The purpose of this paper is twofold: (a) to provide a complete self-contained exposition of estimating life tables with covariates through the use of hazards models, and (b) to illustrate this technique with a substantive analysis of child mortality in Sri Lanka, thereby demonstrating that World Fertility Survey data are a valuable source for the study of child mortality. We show that life tables with covariates can be easily estimated with standard computer packages designed for analysis of contingency tables. The substantive analysis confirms and supplements an earlier study of infant and child mortality in Sri Lanka by Meegama. Those factors found to be strongly associated with mortality are mother's and father's education, time period of birth, urban/rural/estate residence, ethnicity, sex, birth order, age of the mother at the birth, and type of toilet facility.
（a）通过使用风险模型，对含协变量的生命表估计进行完整且自包含的阐述；（b）通过对斯里兰卡儿童死亡率进行实质性分析来说明该技术，从而证明世界生育率调查数据是研究儿童死亡率的宝贵来源。我们表明，使用专为列联表分析设计的标准计算机软件包，可以轻松估计含协变量的生命表。实质性分析证实并补充了梅加马之前对斯里兰卡婴儿和儿童死亡率的研究。那些被发现与死亡率密切相关的因素包括母亲和父亲的教育程度、出生时间、城市/农村/种植园居住地、种族、性别、出生顺序、母亲生育时的年龄以及厕所设施类型。
