【作者】 罗薇；

【导师】 刘建平；

【作者基本信息】 暨南大学 ， 统计学、统计学， 2016， 博士

【摘要】 为了在有限调查资源下满足日益增长的统计需求,将相互关联的住户调查项目进行整合,实施一体化设计,已成为各国住户调查体系改革的重要方向。在理论方面,本文首先按内容特征和内在逻辑关系将住户调查项目分为基本情况类调查、经济特征类调查和社会特征类调查,依据三者的内在逻辑确定调查的主辅关系,进而确定三类调查的调查方法、调查频率和调查方式等设计内容,为住户调查一体化设计奠定理论基础。其次,从内外部衔接两个方向对住户调查一体化设计进行系统研究:从外部衔接的角度对住户调查、人口普查和行政记录进行整体性设计,建立人口普查和行政记录与住户调查抽样框设计以及更新维护的衔接路径,为住户调查一体化设计确立了外部衔接理论基础。从内部衔接的角度,以满足全部住户调查项目需要的主样本设计为切入点,以设计效应和最终类集为工具,构建多阶段抽样下的复杂样本设计的理论分析框架,分析影响复杂样本估计量设计效应的基本要素,解析其对复杂样本估计量设计效应的影响机理及相互作用机制,建立各种要素的设计效应模型及要素组合的设计效应模型,为主样本设计的分析和选择提供有效的方法路径,为住户调查一体化设计建立了内部衔接理论方法基础。在应用方面,在充分利用现行国家调查制度的渠道和机制的基础上,构造出我国住户调查一体化设计的基础框架,对住户调查项目按其调查内容特征和内在逻辑关系进行精简、整合,形成以劳动力调查和住户收支与生活状况调查为核心的住户调查体系,并给出以主样本为主体的我国住户调查的一体化设计思路,其思路和方法可以复制推广到其它调查领域。更多还原

【Abstract】 In order to meet the increasing statistical data requirement under limited resourses,combining and linking relevant household surveys,and implementing integration design,has become an important aspect of the national government statistic reform.In theory,firstly,household surveys are divided into basic topics,economic topics and social topics on the basis of their contents and internal logical relationships.The predominant and subordinate relations of these three topics should be defined by their internal logic to confirm the method,frequency and mode of every single survey,thus to lay a theoretical foundation for the design of integrated household surveys.Secondly,it develops a systematic research on integrated desgin in a system of household surveys,including the external linkages and the internal linkages.It forms integrated design among household surveys,census and administrative records based on the viewpoint of external linkages.It establishs links among of censuses,administrative records,and household surveys’ sampling frame designing and updating.The work will lay the foundations of external linkages theory for integrated desgin in a system of household surveys.In an internal perspective,the master sample,which satisfieds all household surveys’ sample design requirements,provides a starting point.The design effects and the ultimate clusters are introduced as a research instrument to provide a theoretical analysis framework for complex sample under multistage sampling.Within this framework,the complex sample estimator design effects represent a number of components.The mechanism of the effects of these components and their interactions are measured for the design effects of complex sample estimator,and used for constructing the design effect models of components and combination of components.The framework provides an effective path for analyses and options for master sample design.It also forms the base of internal linkages theory for integrated desgin in a system of household surveys.In application,based on the current system of statistics,it constructs a basic framework of the integrate system of household surveys for China.It lists a series of commonly-surveyed topics and analyses the logical relationship among them.Then it formulates a new household surveys system characterizing with two core surveys----the labor force survey and the expenditure and income survey.It also provides a guide for designing the master sample of the integrate system of household surveys for China.The ideas and methods of master sample design can be replicated and generalized to other survey fields.更多还原