双重抽样框下项目无回答插补估计方法研究  

Research on Imputation Estimation Methods for Items Non-response Error in Dual Sampling Frame

作　　者：马金萍 刘小铃 温欢乐 MA Jinping;LIU Xiaoling;WEN Huanle(School of Statistics,Xi’an University of Finance and Economics,Xi’an 710100,China)

机构地区：[1]西安财经大学统计学院,陕西西安710100

出　　处：《统计与信息论坛》2024年第5期3-15,共13页Journal of Statistics and Information

基　　金：国家社会科学基金重大项目“国家治理能力现代化的测度理论、方法及进展评价研究”(21&ZD147);国家社会科学基金项目“新型农村集体经济富民效应测度与稳定性评价”(23BTJ024);国家自然科学基金项目“基于分类识别与随机动力学的细胞内粒子非正常扩散研究”(12101473)。

摘　　要：抽样调查工作中无回答情形不可避免,双重抽样框下亦如此,因此需要对双重抽样框下抽样调查项目无回答造成的估计量偏差进行纠偏校正。首先通过二重抽样获取辅助变量的信息,使用其构造比率估计量与比率型指数估计量的组合估计量对双重抽样框下抽样调查中项目无回答数据进行插补,得到对应各部分子总体的均值估计,再用Hartley估计量的形式对总体总值进行估计。通过计算估计量偏差、均方误差及最优权重系数,对比相同条件下完全回答时同类型组合估计量均方误差的相对精度损失与使用单一比率型指数估计量的相对精度损失,随机模拟结果显示损失率较低,插补方法有效。选择合适的辅助变量构造比率估计量和比率型指数估计量的组合估计量做插补值,更充分利用辅助变量和已回答研究变量信息,基于提出的组合估计量于抽样调查工作具有一定的应用价值。Non-response error is inevitable in sample surveys,even in dual sampling frame.In this paper,the bias of the estimator caused by the items non-response error in sample surveys in dual sampling frame has been corrected.First,the information of auxiliary variables is obtained through two-phase sampling,and the combined estimator of the constructed ratio estimator and the exponential ratio estimator is used to impute the items non-response data in the sampling survey in dual sampling frames,and the mean estimates of the corresponding sub-populations are obtained.The population value is estimated using the form of the Hartley estimator.By calculating estimator bias,mean square error and optimal weight coefficient,and comparing with the relative accuracy loss of the mean square error of the same type of combined estimator and the relative accuracy loss of the single ratio-type exponential estimator under the same condition,the results of random simulation show that the loss rate is low and the imputation method is effective.Since the appropriate auxiliary variables are selected to construct the combined estimator of the ratio estimator and the exponential ratio estimator as the imputation values,the information of the auxiliary variables and the answered research variables is more fully utilized,the proposed combined estimator has certain application value in sampling survey work.

关 键 词：双重抽样框 项目无回答 辅助变量 组合估计量 

分 类 号：C811[社会学—统计学]