Thread by @datepsych

New study on male dating app profiles and matches:

"Only one profile out of 100 was liked by more than 80% of women."

This is an effect that I have described in the past. Although we see high inter-rater agreement on attractiveness, agreement of individual pairs in a sample can be much lower. In other words, there can be high disagreement even for what an attractive face is.

Here are rejects and accepts per profile.

Age predicted attractiveness ratings and swipes for men: men too much older, or too much younger, than the women were penalized.

Pretty consistent with past research indicating that women don't like large age gaps.

Profiles without clear facial photos were also liked less.

Here are accepts/rejects by participant swiping. Also some variation here, with some women being much more selective than others.

The bottom 10% of rejected male profiles - profiles where faces could not be seen clearly.

Important in naturalistic dating app data - lots of profiles get rejected because the photos are really bad, etc, not because person is extremely ugly.

A point in this paper is that it's easier to predict the profiles that get rejected than it is to predict the ones that are accepted.

Asking participants why they swiped on who they did: attractiveness at the top, and personality in the model.

If you are skeptical that "personality" matters at all on Tinder, app research has consistently shown profiles with text descriptions get many times more matches.

They did some brain imaging with an fNIRS during a swiping on profiles task, but actually included no analysis at all of this. They just included the images they created and a short paragraph.

Unfortunately not very informative.

In the introduction section, they didn't explain why they did this or explain why they looked in that area. No statistics comparing areas of activation.

Probably looking in the wrong place. At least for facial attractiveness, activation is usually associated with the orbital frontal cortex and ventromedial prefrontal cortex.

vmPFC also implicated in the selection of attractive and unattractive faces in behavioral tasks.

fNIRS has pretty low resolution and depth, it may not be ideal for something like this. It has a penetration of about 3cm so you're limited to the cortical surface.

You can see where it is placed and that it's not going to reach some areas of interest.

The behavioral data is pretty consistent with past research on dating apps however.

Women are selective in this environment. Matches aren't evenly distributed across male profiles. Big age gaps usually don't help. Attractiveness reported as a top factor for selection.