Is normie approach "hypothesis testing and everything is always normally distributed and every model is linear" dead now that even books and not just journal articles call those prophets false?