本文探讨了交叉试验设计资料的两种等效性检验方法，即双单侧t检验和可信区间法。
Two one-side test and confidence interval method can be used to test the equivalence.
方法根据中心极限定理和连续性校正原理，给出一次近似法和校正一次近似法计算总体率可信区间的公式。
Methods Based on central limit theorem and continuity correction principal , the formula of once approximate method and its continuity correction formula were derived.
目的在对现有总体率可信区间计算方法优缺点进行评价的基础上，研究计算总体率可信区间的一次近似法。
Objective To put forward the once approximate method for computing binomial confidence interval after present methods were evaluated and reviewed.
方法在双正态模型基础下应用ML估计方法计算ROC曲线下面积，正态近似法估计面积的可信区间及假设检验。
Methods On the basis of binormal model, maximum likelihood estimation was applied to areas calculation, and approximately normal method to confidence interval estimation and hypothesis test.
根据资料一致性检验，采用随机效应模型(D- L法)计算合并相对危险度(RR)及其9%的可信区间(9%CI)。
Homogeneity showed that random effect model should be selected to calculate the pooling relative risk (rr) and its corresponding9%confide nce interval (9%ci).
Cox回归分析法显示，房颤患者死于栓塞的风险更高（调整后危险比， 4.33 95 ％可信区间为 1.78至10.52 ） ，而不是讲死于其他原因。
With Cox regression patients with AF were shown to be at higher risk of dying from embolism (adjusted hazard ratio 4.33 95 confidence interval 1.78 to 10.52) but not from other causes.
Cox回归分析法显示，房颤患者死于栓塞的风险更高（调整后危险比， 4.33 95 ％可信区间为 1.78至10.52 ） ，而不是讲死于其他原因。
With Cox regression patients with AF were shown to be at higher risk of dying from embolism (adjusted hazard ratio 4.33 95 confidence interval 1.78 to 10.52) but not from other causes.
应用推荐