Having looked at the abstract of the conference presentation, and his further and more detailed particulars, I sent Lars three comments on 12th October, and uppermost in my mind was whether the experimental set up was a fair match with real interactions.
Are there any methodological reasons to doubt the applicability of the results?
“Make me laugh” seems a very fair test of the “g is sexy” hypothesis, so your experiment seems a strong refutation.
However, the German abitur examination is at quite a high level, so if two thirds of them have achieve that, could there be a restriction of range which would reduce the intelligence effect?
Of course you can always design a better study in some way, but I think we made a substantial effort to get exactly at the Mating Mind hypothesis. I discussed this study quite intensely with Geoffrey Miller (who is a friend of mine), and he agrees that it is getting at what he originally proposed.
German Abitur rates in the population are close to 50% nowadays. So yes, the sample is not completely representative, but we made a good effort to get other guys than the standard students into the lab. I have no doubt that very low IQ is unattractive, and those guys are missing from our sample. But I think we capture the normal range decently well. It is close to impossible to get a completely representative sample into a psychology video lab for an extensive study, unless you can pay them a lot. None of the men studied psychology, by the way.
All the female raters were students, most of them psychology students. Psychology admission in Germany is highly dependent on very good grades, so the female raters were most likely above average and above the male sample in IQ. Crucially this means that assortative mating for intelligence did not bias our results, as this would have only strengthened any tendency for high IQ to increase attraction, which we did not find.
Just saw your blog: "Women were unacquainted with men" did not mean they were virgins, but that they did not personally know any of the men they rated! We did not assess their sexual history, but given their demographics it is safe to assume that most if not all of them had sexual and dating experience.
One unexpected problem with this sample is that g correlated negatively with self-reported Extraversion, which it usually doesn't. This is likely a sampling bias: The extraverted non-students were more likely to come to our lab. All we could do after the fact was statistically controlling for Extraversion.
Comment: I think this covers my questions very well. The procedures were accepted as a fair test of the “bright men are attractive” hypothesis, of which “make me laugh” is a crucial component, and the qualification levels of the target men are not all that much higher than the general population.
I think it also covers most of the other comments, but over to you on that score.