COMPETING HAZARD MODELS OF MORTALITY AND FERTILITY

本项目的目的是开发和应用竞争危险模型 代表整个生命期内死亡率的年龄模式， 上次分娩后的生存率。 死亡率模型 研究的是一个五参数模型取自生态 文学 奇偶生存率模型基于Erlangian Gamma分布（Gamma distribution）。 单纯形法 人口统计数据拟合风险函数的最小二乘法将是 开发的目的是减少应用这些模型所需的计算机时间， 以增加拟合的准确性。 准确性和统计 然后将评估小样本量的两种模型的稳定性 使用模拟数据。 拟进行的申请包括： 1.确定在大范围内表示的参数值的范围 世界范围内的生命表样本，并建立一个生命模型系统 基于该变化范围的表格， 2.死因结构与死亡关系的探讨 和死亡率的年龄模式，使用全世界的样本， 递减的生命表， 3.对北美哈特派人口的分析，以确定 婴儿死亡率对产次间隔的影响， 近亲繁殖的程度。

The aims of this project are to develop and apply competing hazard models representing the age pattern of mortality throughout life, and parity survivorship after a previous birth. The mortality model to be investigated is a five-parameter model taken from the ecological literature. The parity survivorship model is based on the Erlangian distribution (a special case of the Gamma distribution). Simplex methods of fitting hazard functions to demographic data by least squares will be developed to reduce the computer time necessary to apply these models and to increase the accuracy of the fit. The accuracy and statistical stability of both models with small sample sizes will then be evaluated using simulated data. Applications to be undertaken include: 1. Determination of the range of parameter values represented in a large worldwide sample of life tables, and development of a system of model life tables based on this range of variation, 2. An investigation of the relationship between cause of death structure and the age pattern of mortality, using a worldwide sample of full and decremented life tables, 3. An analysis of the North American Hutterite population to determine the influence of infant mortality on parity intervals, and the influence of level of inbreeding on demographic rates.