Thread by @datepsych

Here are the results of the Gigachad survey of facial attractiveness many of you participated in.

Also a review of literature on interrater agreement on facial attractiveness and facial masculine dimorphism. 🧵

datepsychology.com/women-dont-fin…

This is a long article, but I will try to summarize it in a short thread here.

First, survey results.

Women found the Gigachad to be below average. They rated him as less attractive than men did.

Gigachad was also a polarizing face. Few men or women rated him as average. Most women clustered around a very low rating, while most men clustered around high ratings.

Only about 25% of women rated him in the Chad zone at all!

Next I reviewed interrater agreement on facial attractiveness in past research. Basically, is attractiveness subjective or universal - how much do people agree on what an attractive face is?

For mean ratings, correlations are actually pretty high.

But as we can see with the Gigachad example, average ratings can be misleading.

Between-individual correlations are lower.

One example here to show you how much any two raters may disagree on a face:

In Hônekopp above we can also see the degree of shared vs private variance in facial attractiveness ratings. Shared and private account for about equal amounts of variance in ratings.

So beauty is neither subjective nor universal. It is some other third thing.

In Bronstad et al. we can see an assortative effect. Friends, spouses, and siblings agree more on what attractive faces are than strangers.

But since spouses agree the most, Bronstad said a genetic basis for perceptions of facial attractiveness was not supported.

A twin study here. Low to moderate heritability or genetic contribution to perceptions of facial attractiveness. Most of variance was explained by non-shared environment.

Here is a chart from Lavan et al. that, like the Gigachad results, can help you to visualize how much individual raters can vary facial attractiveness even when mean ratings are close.

Test-retest for facial attractiveness also was not great. In other words, people may rate a face highly one day, less highly the next day.

Fairly stable, but not as much as I would have thought.

The tends to be cross-cultural agreement on what faces are attractive. However, that agreement also tends to be lower than within-culture agreement.

Pretty consistent with the large role of non-shared environment, as well as with a non-sejective basis for attractiveness.

In this article I cover a lot of research on masculine facial that morphism and attractiveness - basically, are masculine faces more attractive?

Research is mixed, leaning towards "no" imo.

With some important nuance.

First, a great deal of research on facial dimorphism and attractiveness uses computer edited faces. It seems like most of the research that finds more masculine faces are less attractive uses this methodology. Thus, this may be a confound.

When unedited faces that are rated as more masculine are used, women do actually tend to find these more attractive.

This was a fun study: when women were able to edit the faces themselves, like a slider in a video game creating a character, they turned the masculinity of those faces way down. Big effect here.

It is worth adding as well that faces perceived to be more masculine by women may not actually be the most dimorphic faces either.

There seems to be a disconnect between faces that are rated as more masculine and faces that are objectively more dimorphic.

So, when women prefer more masculine faces, or rate faces as more masculine, they may not actually be faces that look like the Gigachad.

I looked at a lot of cross-cultural preferences for facial masculinity and dimorphism as well.

This may be even more culturally or context dependent than facial attractiveness.

Lots of mixed results here as well. For example, here we see two studies that found opposite trends: more masculine faces preferred in high HDI countries, or less masculine faces preferred in high HDI countries?

Final part here: the Gigachad had some stubble and facial hair is a male dimorphic trait, so I looked at a lot of research on attractiveness and facial hair.

I was surprised to find how consistent female preferences for facial hair were.

This is something that I expected to read more mixed results on, but research seemed to be consistent that women prefer varying degrees of facial hair over clean shaven faces.

It didn't matter if it was stubble, a light beard, a heavy beard, etc - one or more of these usually beat clean shaven faces across all of the studies.

On Likert 5 and 7 point scales, facial hair raised facial attractiveness by about .5 to 1 point.

One example chart:

There is a section in the article about the neuroscience of facial attractiveness.

Summing this up crudely - more attractive faces do stimulate brain responses more. They do so differently for men and women as well.

A few misc research results here:

Gay men are more likely to prefer heavily masculinized faces (like the Gigachad).

This is something I didn't control for, so it may explain the much higher male preference perhaps.

Men in general are also more likely to rate male faces higher than women are. This may also explain higher male ratings for the Gigachad in my survey.

Perhaps a big take-home point here is that mean interrater agreement on facial attractiveness is high, but can be misleading due to the high variance in how individuals perceive a given face.

So, even if you receive a low rating there may still be a very large cohort of men or women who find you attractive.

At the end of the day, you don't need most people to find you attractive anyway. Just one person.

Potentially large disagreement in how men and women rate faces is also worth keeping in mind.

Whole cottage industry of men paying other men for facial ratings.

Lower between-judge correlations male/female differences in rating should make you cautious.