Question

A friend of yours estimates a regression and finds an interesting result that has a p-value below .05 (p = .032). However, they have no idea what a p-value is and only know that a p-value below .05 is "good." How would you define a p-value? Pretend your friend knows nothing about statistics other than how to run a regression.

Answer 1

Answer:

If we want to define the p-value in a non-statistical way, then we can say that the p-value is the specific significance level at which we reject the null hypothesis. This means, the p-value is the boundary for significance level for proving or disproving the claim. For example, if you have p-value 0.032, then this is your standard significance level for this test. It is nothing but the probability of occurrence of given claim or event. This is the smallest level of significance at which the null hypothesis would be rejected. For the significance levels smaller than this p-value, we cannot reject the null hypothesis.