times in this forum, whose participants are quite highly educated. It's not that obvious at all...people without intensive training in this field can be easily deceived by their intuition...

It's not a childish topic, actually, it's a topic that many statisticians and econometricians are doing research on. Andrew Gelman, a statistician in Columbia, devoted many many many blog posts on this topic. James Heckman got his Nobel Prize partly because of the research on it.

While statistics can be highly technical, one of my basic criteria on whether someone has statistical "brain" is to see whether he/she is aware of this point. I think that's the single most important thing in statistics. With it, you are a master of statistics, while without it, a slave (oh, who said "there are 3 kinds of lies: lies, damn lies and statistics"?)