In my opinion, shooting a box test is a complete waste of ammo. Unless you have 100% perfect conditions, a rifle with amazing REPEATABLE accuracy, a concrete bench and BR rests blah blah blah the shooter/rifle will introduce WAY more error into the test than the scope will. Therefore the test will be completely inconclusive.
Best (only) way to do it is to mount the scope to something solid that just won’t move i.e. concrete bench and set a test chart at exactly 100 yards/metres depending on your click values. Look through the scope and dial up across as much as you like through the whole adjustment range. Ensure that your parallax is set properly before the test also.
Very few people are set up to do this correctly.
As BB mentioned backlash will be evident in a lot if not most scopes so going past your zero and then coming back to it from the same direction can help.
The ONLY 2 scopes that I have personally tested that were perfect are my S&B and my NF.
Bookmarks