3 ms·

N=20

• tetris11 4y ago
  Effects still appear to be significant in those studied
  • LTmufufu 4y ago
    You can find some statistical power with any number of observations. Doesn't mean results can be generalized, which is what the article wants to do. 20 extremely low. At that level it is more reasonable to ask why anyone would assume there to be no differences between two groups of ten people.
    • tetris11 4y ago
      when the effect between 10 people is small, agreed, the results aren't significant. But if there are extremely large effect differences, one can't help but raise an eyebrow, even if the n is low