I have 2x2x2 table. I'm using repeated G-tests of independece of 2x2 by 2. I did not apply any correction. My G values are: Sample1 G=1.78 df=1 Sample2 G=3.43 df=1 Total G=G1+G2=5.22 df=1+1=2 Pooled G=5.45 df=1
Now, beacause the Total G is less than the Pooled G I can't get the Heterogeneity G value.
The 2x2 tables are: s1=c(12,1,18,0) s2=c(27,3,22,0)
My attitude is to be like water flowing in a stream. If there is a block, just go around it. It seems that a straight forward application of a given statistical method does not work in your case. Why not start afresh? there are some problems with your data. Each 2x2 table has a zero. yet another count is just 1 (or 3). Not a nice situation! if you tell me more about the actual problem and questions you seek answers for, perhaps some alternative approach can be found.
- Badges
- Science topic