摘要
[目的]对基于二项分布的小样本总体率可信区间估计方法—Clopper-Pearson提出的精确法进行校正,以减少精确法的保守性,提高可信区间的精密度。[方法]运用SAS软件,编制二项分布的Monte Carlo模拟抽样程序,通过计算95%可信区间的实际可信度寻找合适的校正系数k。[结果]校正系数k=0.5时校正法所估计的95%可信区间的实际可信度比精确法更接近期望可信度(95%),其可信区间宽度比精确法更窄;而当样本含量小于15时,k=0.6的校正结果最理想。[结论]校正法可以减少精确法的保守性,提高可信区间的精密度。
[Objective] To adjust the exact method that based directly on the binomial distribution proposed by Clopper-Pearson for constructing confidence interval in order to decrease its conservation and to increase its accuracy. [Methods] The actual confidcnce degrees of exact method and four adjusted coefficients （k = 0.4, 0.5, 0.6, 0.7） were eompared using the teehnology of Monte Carlo simulation sampling. [ Results] The actual confidence degree of 95% confidence interval calculated by adjusted coefficient k = 0.5 was closer to the imposed confidence degree （95%） than that of exact method, and width of the confidence interval was smaller. When sample size is less than 15, the best adjusted result of k = 0.6 was expected. [ Conclusion] The adjusted method is better than exact method for eonstructing small sample sizes binomial proportion confidence interval.
出处
《现代预防医学》
CAS
北大核心
2007年第13期2472-2474,2476,共4页
Modern Preventive Medicine
关键词
可信区间
总体率
二项分布
Confidence Intervals Population Proportion
Binomial Distribution
