摘要
文章阐述了如何用样本出生性别比估计或推断总体出生性别比的问题,以及如何利用两样本性别比来检验两个总体性别比之间在统计上是否存在差异。文中给出了出生性别比的标准误计算公式,并计算了在不同样本出生性别比水平下、不同样本规模的情况下标准误的估计,以及在95%置信度下的置信区间的估计;同时还利用二维交叉表的卡方检验和OddsRatio的卡方检验对两样本所反映的总体出生性别比差异的显著性水平进行了检验。文章得出的结论是,规模比较小的样本(出生人口少于3000人)和人口比较少的总体(总人口少于30万人),均不易于单独计算出生性别比。
This article explores how to use sample data to estimate a population’s sex ratio at birth and how to test the statistical difference between the sex ratios at birth of two populations by using the sex ratios at birth of two samples.The formula for calculating the standard error of sex ratio at birth is given,the estimates of standard error are given in different level of sex ratios at birth and different sample magnitude,and 95% confidence intervals are calculated.In addition,Chi-square test for two-by-two tables and Chi-square test for Odds ratio are used for testing the significance of difference between two sex ratios at birth in populations.The conclusion is that it is not appropriate to use a small scaled sample(less than 3000)or a small number of population in an area(300 thousand in total population)in statistical inference for sex ratio at birth.
出处
《中国人口科学》
CSSCI
北大核心
2006年第6期30-35,共6页
Chinese Journal of Population Science
