Emil O W Kirkegaard

Cool! That even includes interest in casual sex, which I thought would be a pretty good indicator.

Of course, these participants are selected for being on OKCupid, a left-wing dating site, so somewhat range restricted which probably reduces sex differences a bit -- but still. Surprised it's only 72 even with RF. I could try more, but already took me hours to fit this on laptop

Isn't a dating site very sensitive to response bias? You're answering questions in front of prospective future partners.

There's also the problem of knowing thyself. I've seen plenty of people claim they're empathetic when they're not, or that they try to avoid drama when they're one of the most drama-ridden people I know, etc.

Yes, self-reports are always problematic but I'd guess that self-reports on dating sites are especially prone to bias

I can see the errors working towards providing separation though. If there is a difference between the ideal man and woman than each gender will try and mark those answers. Men might be willing to say they take risks (smoke/drink), while women demure on that.

@KirkegaardEmil, this data is available online?

Even if it's not respondents are more likely to give responses that would match them with the kind of partner they seek. Those will be different on a site like OkCupid compared to a more hookup oriented site as Tinder.