Phil Batterham, an epidemiologist at Australian National University in Canberra, wonders what aspects of intelligence made soldiers more likely to die in the war. "One could hypothesise that the association between greater intelligence and higher war-related mortality might be driven by the more crystallised verbal abilities, leading to greater leadership roles," as opposed to other forms of intelligence, he says.

On the face of it there seems to be a flaw with sample size re Scots several hundred vs several thousand. Can you draw reliable conclusions from that comparison?

not to mention 3 IQ points isn't really much of a difference. But I would teen to agree with the leadership hypothesis.

We are all left to ponder these and many points until of course we read the study. I for one am curious about any and all confounds the author(s) discussed in the article. Though effect size and N are important, let's think about how it was that the investigators arrived at their H0 and H1.

No wonder the Bush family has been around so long.