Comment by nonbel

A model of whatever process you think generated the data.

EDIT:

I guess I should say that the concept of testing a "null model" without interpreting the fit relative to other models is wrong to begin with. You need to use Bayes' rule and determine:

  p(H[0]|D) = p(H[0])p(D|H[0])/sum(p(H[0:n])*p(D|H[0:n]))

Lots of stuff wrong with what has been standard stats for the last 70 years, it literally amounts to stringing together a bunch of fallacies and makes no sense at all.