P-Value and hypothesis testing theory
In: 2
Null hypotheses, Type I and II errors—these are beyond a 5-year old, I think. So:
You want to compare two things: drug vs. placebo, radiation vs. chemotherapy, standard vs. underhand foul shooting, etc. You run an experiment and find a difference. Problem is, the difference could be real or just due to random factors. You ask a statistician, and they tell you “The difference is real, p < .05.” This means there’s less than a 5% chance that random factors caused the difference. You now have evidence (not proof, just evidence) that your treatment worked.
Latest Answers