Demographic Analysis of Longitudinal Historical Data

描述(由申请者提供)：该项目将为从事不同人口统计领域工作的学生和研究人员提供非常有价值的历史人口统计技术方面的专门培训，以分析纵向数据。这个项目的理由很简单：历史人口学对人口研究的理论、方法和实践做出了重要贡献，特别是在纵向数据的使用方面。历史人口学家目前正在对生育率、死亡率、家庭制度、老龄化和移民等主流人口学研究做出重要贡献。事实上，现有和在建数据库的规模、范围以及时间和地理覆盖范围都是史无前例的。由于历史数据往往是纵向的和多层次的，它们提出了微妙的方法问题。有意义的分析往往需要专门的方法，如家庭重组和反投影，这是历史研究所独有的。由于它们以人口统计学的基本原理为基础，因此接受过这些方法培训的学生既能为历史研究做好准备，又能更好地使用复杂的当代资料来源。历史数据可以成为分析人口统计过程的完美模型。观察到的协变量的数量通常是有限的，历史人口学家在创造性地使用纵向和谱系信息来构建上下文和随时间变化的协变量方面表现出色。学生学习的纵向分析技术将提供一个路线图，用于任何具有时间维度的数据集，包括通过NIH基金收集的许多大型当代数据集。该计划将提供正式课程和与活跃的研究人员实践经验的机会。将向学生介绍当前研究前沿的数据集和先进的统计技术。我们还致力于保护和存档早期研究中使用的一些经典数据集，并使它们可用于使用最现代的方法进行分析。

DESCRIPTION (provided by applicant): This project will provide extremely valuable specialized training in historical demographic techniques for analyzing longitudinal data to students and researchers working in a variety of demographic sub-fields. The rational for the project is simple: Historical demography has a long history of important contributions to the theory, methods, and practice of population studies, especially in the use of longitudinal data. Historical demographers are currently making important contributions to mainstream demographic research in fertility, mortality, family systems, aging, and migration. Indeed, the size, scope, and temporal and geographic coverage of databases currently available and under construction are unprecedented. Since historical data are often longitudinal and multi-level, they raise subtle methodological problems. Meaningful analysis often requires specialized methodologies, such as family reconstitution and back projection that are unique to historical research. Since they are based on fundamental principles of demographic theory, students trained in these methods are both prepared for historical research and better able to use complex contemporary sources. Historical data can be a perfect model for analysis of demographic processes. The number of observed covariates is usually limited, and historical demographers have excelled in creatively using longitudinal and genealogical information to construct contextual and time-varying covariates. The longitudinal analysis techniques students learn will provide a roadmap for use with any data set with a time dimension, including many large contemporary data sets collected through NIH funding. This program will offer both formal courses and opportunities for practical experience with active researchers. Students will be introduced to data sets and advanced statistical techniques at the forefront of current research. We also aim to secure and archive some of the classic data sets used in earlier research and to make them available for analysis with the most modern methods.