Dear All
Consider three testing methods A , B and C for test of hypothesis
( for example , test of difference two population means under
heteroscedasticity). by use of simulation, one type error of
method A is 0.33 and for method B is 0.42 and for method C is 0.57
While power of test for method A is 0.763 and for method B is 0.822
and for C is 0.911 .
which one of three methods is robustter ?
(note: significance level is .05)
Thanks
Zahedian
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|