3 ms·
Running the experiment until you have a specific pre-determined number of observations is okay. However, the deceptively similar scheme of running it until the
by aliceryhl 3y ago
Running the experiment until you have a specific pre-determined number of observations is okay.
However, the deceptively similar scheme of running it until the results are statistical significant is not okay!
- mreezie 3y agoIf you want statistical significance of 1/20 and you check 20 times... you are likely to find it.