3 ms·
> How long should the trial run before you decide that the null hypothesis is true? As long as your experimental design says you should. You don't have a scien
by erikpukinskis 9y ago
> How long should the trial run before you decide that the null hypothesis is true?
As long as your experimental design says you should. You don't have a scientific outcome until the trial is over, so stopping makes no sense on a scientific basis.
- carbocation 9y agoOf course. The problem is that trials when your effect size is small are expensive. A trial looking at longevity outcomes for young healthy people would require decades of follow-up. The expense would be enormous, and the results would come generations later. Really have to hope you've asked the right questions.