Back to Community
Jarque-Bera Test and p-values


Re the Instability of Estimates lecture, I'm having a hard time parsing the meaning of the p-value returned by running the Jarque-Bera test. In the lecture notebook, this is stated:

Sure enough the value is < 0.05 and we say that X is not normal

And yet, in the answers to the exercises notebook, this code is run:

if Xp < 0.05:  
    print 'The distribution is likely normal.'  

These aren't consistent, unless I'm missing something? The first makes more sense to me--given the null hypothesis that X came from a normal distribution, then a significant p-value (less than 0.05) suggests that X likely did not come from a normal distribution.

Can someone please explain if I'm misunderstanding?