c**********e
发帖数: 2007 | 1 If the two ratios are from two group of subjects,
it would be easy.
Now the two ratios are from the same group of subjects.
One before treatment (only one treatment), and one
after treatment. For example, before treatment,
55% (55 out of 100) of patients have high blood pressure.
After treatment, 50% (50 out of 100) have high blood
pressure. Are the two ratios significant?
Thanks a lot! |
p********a
发帖数: 5352 | 2 TTEST不就OK了？WHY do you care if they are in the same group? |
s*r
发帖数: 2757 | |
c**********e
发帖数: 2007 | 4 Great. Thanks.
【在 s*r 的大作中提到】
: mcnemar test
|
o****o
发帖数: 8077 | 5 related response from matched-pairs
【在 p********a 的大作中提到】
: TTEST不就OK了？WHY do you care if they are in the same group?
|
p********a
发帖数: 5352 | 6 完蛋了，俺一直用TTEST。。不过我的GROUP一边大，一边小，大的包括小的
【在 o****o 的大作中提到】
: related response from matched-pairs
|
l***a
发帖数: 12410 | 7 then what's way to go
【在 p********a 的大作中提到】
: 完蛋了，俺一直用TTEST。。不过我的GROUP一边大，一边小，大的包括小的
|
c**********e
发帖数: 2007 | 8 Independent if from different group. Otherwise not independent.
【在 p********a 的大作中提到】
: TTEST不就OK了？WHY do you care if they are in the same group?
|
l*****i
发帖数: 43 | 9 paired_t test, because this is before and after test, typical paired_t test. |
