That was the point of my post. You _can't_ say this or that based on a small sample size; there will almost always be a counter example in sociology (and in most sciences).