If your generalization is an accident of population samples, and doesn't apply when you control for that, I'm not sure how you think it's useful. If you think

If your generalization is an accident of population samples, and doesn't apply when you control for that, I'm not sure how you think it's useful. If you think that anti-PC people are simply scared of being called assholes, you missed the time when someone got fired making a joke about dongles. And the time when someone else got fired for saying that Donald Sterling shouldn't have been fired for things that he said in his own home. And the time when Bahar Mustafa was arrested, although I acknowledge that a subset of anti-PC people seemed to celebrate that. And the time when a cake chain got successfully sued for not wanting to bake a cake they disagreed with. Just off the top of my head.