To address the statistical concerns, the assumption in using a Audette test is your rifle has to be accurate enough to give you high confidence that the elevation differences are load related and not random error (rifle with poor accuracy). A poorly shooting rifle gives too much random noise to use an Audette test.