5 ms·

I don't get it. Isn't "every MLB player who agreed to play" too small a sample size to be useful?

• im3w1l 6y ago
  Short answer: No. This question comes up a lot. Someone should really write a blog post about how big sample size you need, and the dangers of using a biased sample (fit men), and spam it to death here.
  • hcknwscommenter 6y ago
    I think the "fit men" can be an issue when you are testing whether a drug works to reduce severity (treat) an infection. However, the question of whether the vaccine actually prevents infection should be relevant regardless of the level of fitness. You should probably only vaccinate half of them and when inevitably an unvaccinated person gets it, you should have very good data about protection two weeks later. This is all just a thought experiment of course (unless there are volunteers . . .)