SAS中文论坛

标题: kappa系数和McNemar’s检验 [打印本页]

作者: shiyiming 时间: 2011-11-23 10:32
标题: kappa系数和McNemar’s检验
现在数据中有A、B两个变量，A为组别，B为评价结果，数据形式例如
A  B
1  1
1  2
1  2
2  3
2  4
2  1
现在要做1组与2组评价的交叉比较
............2组
....... .....无法评估.......差........一般........好
1组
无法评估.....A
差...........................B
一般....................................C
好....................................................D

并且要算一致率（（A+B+C+D）/总人数）及其95%CI
kappa系数及其95%CI
McNemar’s检验的统计量以及P值

请问一下，其中一致率的95%CI
kappa系数及其95%CI
McNemar’s检验的统计量以及P值
这些统计量分别应该如何编程计算？

作者: shiyiming 时间: 2011-11-25 09:22
标题: Re: kappa系数和McNemar’s检验
McNemar's test is on 2 by 2 table (at least in SAS so far); so it is not computable in your study.
the Kappa test is on n by n table for 2 raters; that is, you got to make up those missing grades for raters. I assume in your case that the two rates are rating 3 things.
No idea on that rate you have mentioned, a +b+c+d, what is that? the diagonal cells?

JingJu

欢迎光临 SAS中文论坛 (https://www.mysas.net/forum/)