c**********e
发帖数: 2007 | 1 Suppose that I have a group of subjects, which has
scores 0 (normal), 1 (moderate), or 2 (severe) for
a certain disease at baseline.
At the end of treatments, the subjects have another
scores 0 (normal), 1 (moderate), or 2 (severe).
How to test if the patients have improvement?
Thanks a lot. | D******n
发帖数: 2836 | 2 check "McNemar's Test"
【在 c**********e 的大作中提到】
: Suppose that I have a group of subjects, which has
: scores 0 (normal), 1 (moderate), or 2 (severe) for
: a certain disease at baseline.
: At the end of treatments, the subjects have another
: scores 0 (normal), 1 (moderate), or 2 (severe).
: How to test if the patients have improvement?
: Thanks a lot.
| s*r
发帖数: 2757 | 3 check agresti's book
it seems the author does not suggest mcnemar's test can be extended to 3 by
3 table, instead logistic regression for correlated response was suggested |
|
