A survey in Norway shows that the oldest brothers' IQ is higher by about two points relatively to their younger brothers. The article proposes some sociological theories based on interactions between the brothers. Your humble correspondent thinks that this theory is mostly bogus.
In fact, it's also known that younger brothers are more likely to be gays. Their brains are correspondingly less male in average. There is a good reason - a likely mechanism - behind these facts: mother's womb gradually develops a certain kind of immunity against the male foreign intruder that acts on the younger brother's foetus.
Note that this action of hormones and the antibodies requires time. That's one of possible explanations why similar considerations can't influence the very sex of the child. Indeed, as far as I know, there exists no indication that younger siblings are more likely to be girls or more likely to be boys. In fact, the data are good enough that it is known that the existence of older sisters doesn't enhance a man's probability of being a gay, unlike the existence of older brothers.
A lot of facts are already known today and many more may be found in the future. Some of these facts are very sharp and a generic explanation of these observations based on social interactions may be instantly falsified.
Much like crackpots in high-energy physics such as Smolin or Woit, the defenders of social explanations of these effects don't appreciate how difficult and unlikely it is to create a theory that is consistent with all the available data. They think that it is always possible to emit fog and argue that things are complex and theories with arbitrary underlying ideas are hard to be falsified. But in reality, it is easy to falsify almost all proposed theories simply because a lot of non-trivial facts and regularities are known. And that's true both in quantum gravity as well as the science about intelligence.
And that's the memo.