[ravenstine]

The data is interesting, but I see it as poor quality. People are horrendously bad at self assessment on average, and many have "main character" syndrome that biases them towards claiming favorable traits they may lack in practice.

There's a reason you don't ask a woman for advice on dating women, or men on how to date men. Most people, particularly young people, don't actually know what they want. Personality tests aren't worthless, but you're essentially asking people about themselves that way.

A better study should ask one sex their perception of the other. This is imperfect, but I believe personal experience is more useful than self assessment or personality tests. In the end, it's how the sexes actually interact that matters.

[greenonions]

It's true that people don't have any idea about themselves, but I'm very skeptical you would get any better data by asking a different sex.

Just as much as people don't know themselves, they often know even less about others.

[Melatonic]

I think a much better way would be to ask the best friends of people to anonymously rate their friend and how they date. The best friend will be close enough to probably have a good idea. Obviously they like the person if they are friends (so there are biases here as well) but they might be more likely to notice patterns the person is otherwise oblivious to.

[riedel]

Sciences and scientists (quoting the terms used in the OP site) that publish statistics as bar charts in company blog post typically make it just a bit difficult to assess any quality at all. I simply would not even bother trying to interpret the data in its current form.