A scientific conclusion is only as good as the data it is based on.

If the data is sparse/non-existent, incomplete, noisy, or biased, then no matter what the p-value is, the conclusion cannot be strong.

— Alina Chan (@Ayjchan) October 24, 2022