摘要
本文对韩国国内行政区域如市郡区等小地区的失业统计共同差额比进行推定的方法有联合推定量,Woolf推定量和Mantel-Haenszel推定量等。对这个推定量的可信度是通过偏差和平均平方误差的概念来进行比较的。从对京畿地区经济活动人口调查情况看,这个地区内的24个市郡单位行政自治地区的男、女失业率差额比的偏差及平均平方误差是通过本次研究提出的推定程序来推定的。这些推定数的稳定性和效率性可以通过相对偏差和相对平均误差平方根来评价。Woolf推定量或Mantel-Haenszel推定量比联合推定量更为稳定,从其效率性来看三者很相似。
There are different kinds of methods to estimate the odds ratio for unemployment statistics in small areas, namely,the joint estimator,the Woolf estimator and the Mantel-Haenszel estimator.We can compare the reliability of these estimators according to the deviation and MSE.This estimation procedure established by this study has been applied to estimate the deviation and MSE of the odds ratio between the male and female unemployment rate in some small areas.The Woolf estimator or the Mantel-haenszel estimator is more stable than the joint estimator,but all these three estimators are similar to each other from the aspect of efficiency.
出处
《中国流通经济》
CSSCI
北大核心
2005年第2期61-64,共4页
China Business and Market
关键词
小地区
失业率
差额比
推定量
small area
unemployment rate
odds ratio
estimator
