论文部分内容阅读
在临床实验中经常遇到的两个率的等价性检验我们已作过介绍,但在医学科研设计中还会遇到比较两均数是否相近或相等,即称为等价的问题。例如观察某一药物治疗高血压病,我们希望它的效果与标准药物利血平相近或相等,因为新药的付作用小,价格便宜,易于生产等。要解决这类问题,就需要借助于等价性检验,而不能直接利用惯用的假设检验中计算α值的方法来进行,因为这时即使得出 P>0.05,也只能说明“不能认为μ_1(?)μ_2,但并不能得出两者相近或相等的结论。而这点在实
The equivalence tests of two rates often encountered in clinical trials have been introduced. However, in medical research and design, we also encounter the question of whether the two means are similar or equal. For example, to observe a drug for the treatment of hypertension, we hope its effect is similar to or equal to the standard drug reserpine, because the new drug has a small effect, is inexpensive, and is easy to produce. To solve this kind of problem, it is necessary to use the equivalence test instead of directly using the method used to calculate the value of α in the conventional hypothesis test, because even if P>0.05 is obtained, this can only indicate that “μ1 cannot be considered. (?)μ_2, but it does not lead to the conclusion that the two are similar or equal.