It is very easy to forget where the actual difference is located: Imagine for example that in some group, on average, boys are 5% more interested in science, that is a fairly useless piece of information if the spread within both girls and boys is very large. I belive humans (only featherless ones though), on average, are to quick to try to derive answers from to little data: in the example above, knowing that someone is a boy or girl has no practical bearing on whether they are likely to be scientists: One should conclude “not enough data” instead. Saying to ones brain that something “has this effect but only very weakly” very often lends the effect to much weight.
It is very easy to forget where the actual difference is located: Imagine for example that in some group, on average, boys are 5% more interested in science, that is a fairly useless piece of information if the spread within both girls and boys is very large. I belive humans (only featherless ones though), on average, are to quick to try to derive answers from to little data: in the example above, knowing that someone is a boy or girl has no practical bearing on whether they are likely to be scientists: One should conclude “not enough data” instead. Saying to ones brain that something “has this effect but only very weakly” very often lends the effect to much weight.