摘要
中国在1984年确定了国家抽样调查县,这些调查县一直使用至今,已有20多年。利用这些调查县,中国进行了许多关于农业的调查,并且有一些原因一直没有改变这些调查县。近年来,有人对这些调查县的代表性提出疑问,是可以理解的。我们应该解释这些调查县能被保持不变的原因,并能找出改进调查的好办法。本文给出一种方法,就是使用调整系数来解决这些调查县的代表性问题,并且利用第一次中国农业普查的数据来对方法进行模拟,得到了较好的结果。我国2007年正在进行第二次全国农业普查,我们可以应用普查结果来完善抽样调查,提高国家抽样调查县的代表性。我们建议结合面积抽样框进行多样框抽样设计,并可以考虑以县作为中国农业调查的初级抽样单元。
It is more than twenty years for China to establish the state survey counties since 1984. Many surveys about agriculture in those state survey counties have been done and those state survey counties have not been changed for many reasons. It is understandable that many persons question the representativeness of those counties. We should explain the reason that we can keep those counties and find a better method to do the survey. This paper gives a method of using adjustment coefficient to solve the problem about the representativeness of those counties and the simulation of the method with the data of the first agricultural census in China is also given. The simulation result is pretty good. The second agricultural census in China has been holding in 2007, we can apply its result to improve the sampling survey and to raise the representativeness of the state survey counties. We suggest using multiple frames to design agricultural sample survey.
出处
《统计研究》
CSSCI
北大核心
2007年第11期69-73,共5页
Statistical Research
基金
教育部人文社会科学研究重大项目"我国政府统计抽样调查制度的理论方法研究"(O2JAZJD910002)资助
关键词
国家抽样调查县
代表性
系数调整
抽样误差
State survey counties
Representativeness
Coefficient adjustment
Sampling error
