珍珠湾ART

标题: 请教个问题 [打印本页]

作者: 动脑筋 时间: 2008-7-23 22:59
标题: 请教个问题

假设有T种新药(T1, T2, T3 etc.)，针对同一种致命疾病。
将同样病情的病人，随机分成T组(T1， T2， T3 etc.)，每组人数相同(V人)。
每种药在相同测试时间内只能产生两种结果：生/死
每种药在相同测试时间内都有一个成活率百分比 (R1， R2， R3 etc.)
取R值最好的两个组，计算R的差值 Rd

请问：根据V值和T值，Rd至少应该大於多少，才有效？
例如：当V = 50， T=5 的时候如果Rd= R3-R2 = 0.2 可不可以说T3好于T2？
当V= 1000000， T = 3 的时候如果Rd= R3-R2 = 0.2 可不可以说T3好于T2？www.ddhw.com

Conditions:
1) T types of medications, all for the same disease;
2) Randomly assign patients of the same condition to T groups - the number of patents in each group is V;
3) Within the same period of time, each group takes one type of medication which yields only two results: Survive/Die;
4) Within the same period of time, the percentage of people who survive in each group is R;
5)Rd is the difference between the top two R values.

Question: Based on T and V, how to calculate the lowest Rd value to prove one medication is better than the others?
www.ddhw.com

作者: 大清太平 时间: 2008-7-24 21:47
标题: 又考统计学了[:((]

significance tests