S4E15 Flipping Our Hypotheses to Test Equivalence
Quantitude
00:00
Do You Hate the Null Hypothesis?
KG: From a statistical standpoint, there is insufficient empirical evidence to say that my observed difference between means would have been unlikely given the null hypothesis. But if we retain the null hypothesis and even the language that I'm using is intentionally guarded, some textbooks will actually say that we accept thenull hypothesis. That's just incorrect. KG: There really is merit in saying, clumsily, that we failed to reject the null hypothesis - which is not the same as saying there's no difference.
Transcript
Play full episode
