| 摘要: |
| 目的在固定样本抽样框架下结合抽样和检测成本探讨基于混合样本的感染率检验中的混合样本大小和检测次数的确定方法。方法利用反正弦变换和正态近似,在控制两类错误概率的前提下建立混合样本量与混合样本大小之间的函数关系,并通过随机模拟考察近似功效函数的近似程度。结果利用数值方法得到不同检验阈值下最小检测次数的整数解及对应的最优混合样本。结论给定两类错误概率,混合样本的大小与检测次数之间此消彼长,实际工作中应根据样品采集成本与检测成本的相对大小加以权衡。 |
| 关键词: 感染率 临界阈值 混合样本 检验功效 |
| DOI：10.3724/SP.J.1008.2011.01353 |
| 投稿时间：2011-09-19修订日期：2011-10-09 |
| 基金项目: |
|
| Optimal pool size and pooled sample size for hypothesis test of critical threshold of infection rates based on fixed sample size and pooled sampling method |
| SUN Qing-wen1, TENG Hai-ying1, SONG Mao-hai2, FANG Ying1* |
(1. Department of Mathematics and Physics, College of Basic Medical Sciences, Second Military Medical University, Shanghai 200433, China
2. Department of Computer Science, College of Basic Medical Sciences, Second Military Medical University, Shanghai 200433, China
*Corresponding author.) |
| Abstract: |
| ObjectiveTo determine the optimal pool size and pooled sample size for testing whether an infection rate has exceeded the critical level of malaria epidemics using the pool sampling method and fixed sample size approach. MethodsThe function between the pooled sample size and pool size was deduced by using arcsin transformation and normal distribution approximation while controlling the probability of type Ⅰ and type Ⅱ errors. Computer simulation was used to evaluate the approximate power function. ResultsThe optimum pool size and the pooled sample size were obtained for different critical and normal levels of infection rates. ConclusionThe optimal pool size and the pooled sample size are in an inverse relationship for given probability of type Ⅰ and type Ⅱ errors, so in practice we should make an evaluation according to the sampling cost and test cost. |
| Key words: infection rates critical threshold pooled sample test power |
