不同总体量和样本量时如何计算比例的置信区间  

How to Estimate Confidence Intervals of Proportionsfor Different Population Sizes and Sample Sizes

作　　者：吴喜之[1] 

机构地区：[1]中国人民大学统计学院,北京100872

出　　处：《统计与信息论坛》2005年第3期8-10,共3页Journal of Statistics and Information

摘　　要：在总体或者总体子集不大情况下的抽样调查中,往往不易得出合理的关于比例的区间估计。这一类问题在抽样调查实践中已经严重到非说不可的地步。文章讨论了在样本量不大或者(和)在总体不大时估计比例的置信区间时往往忽略的问题,并给出了在不同情况下如何计算置信区间的方法。When the size of population or its subsets is small, a sample survey often cannot provide proper confidence intervals of proportions. In our practice, this kind of problem is too serious to keep silence. This article discuses the problems occurring in estimating confidence intervals when population size and/or sample size is small, which had often been neglected, and provides algorithms of estimating confidence intervals in various situations.

关 键 词：抽样调查 比例的置信区间 二项分布 超几何分布 样本量 总体量 

分 类 号：F224.7[经济管理—国民经济]