摘要
相对风险是流行病学研究中的重要指标之一,它是度量一种暴露因素是否与某病的致病有联系的统计指标.以该指标的数值大小来表明这一暴露因素对某病的发生具有何种影响及影响的大小,体现了暴露与疾病的关联程度.精确地得到相对风险指标的区间估计,对病因推断具有重要意义.但是由于相对风险指标的估计量是两个概率值的估计量的比值,要得到其精确分布一般而言是很困难的,因此已有研究成果大都采用渐近方法估计相对风险的置信区间,这在小样本情况下表现不佳.在二项抽样条件下,对相对风险的点估计、置信区间估计一直被人们所关注.在二项采样下利用鞍点逼近的方法构造相对风险的置信区间,并通过实例与蒙特卡洛模拟,与传统的置信区间构造方法对比,模拟结果显示其优点,尤其是在小样本量条件下估计效果比较好.
The relative risk（RR）between the exposed and non-exposed groups is certainly one of the most important statistical indices to quantify the strength of the association be- tween a given disease and a suspected risk factor in epidemiological studies. To obtain an accurate interval estimation of RR is very important to give the deduction of the cause of disease.But as a ratio of two probabilities,it＇s hard to get its exact distribution.So previous research findings concentrated on the asymptotic methods of interval estimation. Schol- ars have paid attention to the estimation of the RR and its confidence intervals for a long time under the binomial sampling. The article uses the saddlepoint approximation to estimate confidence intervals of the RR and four methods are compared by extensive Monte Carlo simulation. The result shows the advantage of saddlepoint approximation in estimation,especially when the size of the sample is small.
出处
《数学的实践与认识》
CSCD
北大核心
2014年第21期204-217,共14页
Mathematics in Practice and Theory
基金
教育部高等学校博士学科点专项科研基金(20130004110007)
国家自然科学基金(11271368)
国家社会科学基金重点项目(13AZD064)
全国统计科研计划项目(2011LZ031)
北京市哲学社会科学规划项目(12JGB051)
中国人民大学科学研究基金项目(10XNK025)
关键词
相对风险比
二项抽样
区间估计
鞍点逼近
蒙特卡洛模拟
relative risk
binomial sampling
the estimation of confidence intervals
saddlepoint approximation
Monte Carlo simulation
