摘要
目的:对适用于多个率的多重比较方法进行评价和选择。方法:用SAS软件编程模拟计算,以条件及非条件情形下各种方法的Ⅰ型误差、Ⅱ型误差和误差(Ⅰ型+Ⅱ型)值作为评价和选择的指标。结果:给出在所有条件及非条件情形下的Ⅰ型、Ⅱ型误差估计值。结论:给出了控制Ⅰ型误差、Ⅱ型误差或误差3种不同情况下,多个实验组与对照组比较及多组间的两两比较方法的不同选择。
Objective ：To analyze the methods for multiple comparisons of rates. Methods： Simulation experiments were made by SAS Software. Type Ⅰ error, type Ⅱ error estimate of tests were as the indexes for evaluation and selection. Results： The estimation valus of type Ⅰ error and type Ⅱ error were given. Conclusion： How to select the method for multiple comparisons of rates in order to control the type Ⅰ error or type Ⅱ error is introduced.
出处
《南京医科大学学报（自然科学版）》
CAS
CSCD
北大核心
2005年第12期962-965,共4页
Journal of Nanjing Medical University(Natural Sciences)
关键词
蒙特卡洛
多重比较
误差估计值
monte carlo
multiple comparisons
error estimate value
