> Hi, thanks for all that sent me comments !!! But, I am feeling that the
>>question is not solved. One problem is.... if the people returning is the
>same >of the first vaccine application, then we have dependent data here
>and common >chi-square, permutation, binomial, fisher, signal tests, etc.
>cannot be used; >am I right ?
Why not test the change in behaviour between tests?
Thom
|