3 ms·
> P<0.05 doesn't mean 1 in 20 are wrong > It does mean that ... 1 in 20 would be explainable by the null hypothesis. That seems like splitting a pretty fine h
by lisper 9y ago
> P<0.05 doesn't mean 1 in 20 are wrong
> It does mean that ... 1 in 20 would be explainable by the null hypothesis.
That seems like splitting a pretty fine hair to me, particularly since I haven't actually committed to any of the many possible interpretations of the word "wrong".