摘要
本文研究了国内外非抽样误差测度的主要方法,将其分为两类:基于设计的方法和基于模型的方法。文章运用路径分析法重点研究了基于模型的非抽样误差测度方法,包括多指标和单指标统计调查中非抽样误差的测度,并讨论了不同模型的估计与比较。文章总结了国外非抽样误差研究对我国的有益启示,指出要促进中国的非抽样误差研究,使其对中国的统计调查实际有指导意义,就必须在领会国外学者研究非抽样误差测度思想的基础上结合中国的实际情况进行研究和探索。
This paper studies the main techniques of measuring non-sampling error and classifies them into two classes： based on designs and based on models. The paper emphatically studies the latter by the means of path analysis, including the measurement of non-sampling error in the investigation of single index and multi-indices, and discusses estimation and comparison of different models. The paper summarizes some instructive enlightenments for the study on non-sampling error in China, and then point out that we must prompt the research of non-sampling error considering China＇ s situation and referring to foreign scholars＇ techniques in order to make it useful for China＇ statistical investigation.
出处
《统计研究》
CSSCI
北大核心
2010年第7期89-94,共6页
Statistical Research
关键词
非抽样误差
测度
路径分析
Non-Sampling error
Measurement
Path analysis
