Skip to content

Latest commit

 

History

History
4 lines (4 loc) · 468 Bytes

Null_hypothesis_testing.md

File metadata and controls

4 lines (4 loc) · 468 Bytes
  • We assume a null hypothesis that we want to disprove
  • We calculate our test statistic assuming the null is true
  • If we get an extreme value, i.e. one with a small probability of something as large or larger occurring under the null then we have evidence against the null, i.e. we would reject the hull in favour of the alternative.
  • A lot of times resarcher's default to defining "extreme" as something with probability of occuring no more than 5% under the null.