Note from a practicing scientist: if your experiment doesn’t meet any of the evaluating criteria that you set out in advance, then the experiment didn’t work. Appropriate statistics exist if one absolutely needs to change standards after the fact, but post hoc tests are even harder to pass than the usual kind. That accounts for the fact that you can shop around to find the one that best fits the data.
Scientifically speaking, happy-talk surge boosters like Kagan rank somewhere between the disappointing undergraduate assistant whom you mostly try to prevent from injuring himself, and a beaker.