Hi team, I have a quick question on stopping rules in the bayesian world and using custom stopping rules. Is it methodologically sound to use something like
treatment achieved an XX% to beat control within N weeks
to call a winner for an experiment?
Just to add some clarification to the above. Is it actually valid to use an arbitrary stopping rule on the basis of probability to win to identify when to stop an experiment (assuming of course we set that rule before the launch of the experiment).