Dear Helmut,
You're right, procedures like this should be validated!
For the validation of our power procedure, see this OHBM poster: http://users.ugent.be/~jdurnez/ProsPeakPow_1406_OHBM.pdf
Based on a simple model (very few assumptions, no 'new' assumptions compared to signficance testing), we get good results.
Best,
Joke
|