ParentFull threadobastani·It's fine if you run a separate experiment to justify the hypothesis. If you choose the hypothesis after conducting the experiment, then the p-values you obtain for that hypothesis are invalid.View on HN