Women can reinvent, often take different relationship roles, and some guys are into women with low self esteem.
1. Criteria for men and women are very different: tall vs. thin is key. If the profiles list height but not BMI, that would lead to discrepancies.
2. Heterosexual women are more likely to agree on which men is attractive then the reverse (which is a classic problem in dating apps: then attention of fewer women is also more focused)
3. Make-up: unattractive men don’t have that many commonly used options to be more attractive (growing a beard, painful surgery), while women are more willing to spend half an hour to change that. It levels the playing field making the original measure less relevant.
4. When the data was collected:
> assessed when individuals are around 15 years old
My memory of biology class is faint, but I believe that men change less after 15 than women do.
In general, that’s an area that was understudied 25 years ago. I’m happy to read so many references, but couldn’t find my old sociology professor. He was struggling with the impact of height on one’s professional career. He published a lot using a sample of alumni from the other university where he taught (to control for education). Strangely, the impact of height became weaker over the decades… He had so many theories (computers, mostly)—only to realize (after two publications) that it was because the cohort was increasingly feminine. So, he published in succession that:
1. height was key;
2. the gender gap was more than explained by how women were shorter (that paper was not popular);
3. height mattered a lot less for women, therefore the gender gap thing was still a problem, but athleticism was (he had an athletics grade). (That’s how I learned that `log_salary~height*gender`, `log_salary~height+gender` and `log_salary~height:gender` meant completely different things.)
3 bis. strike that: athletics mattered for both, but less than expected; BMI was key for both but mostly women. (That was the example other professors used to teach us about the scissor effect in regression: BMI and height are related.)
4. Height and BMI were proxies for attractiveness that mattered for both but had to be measured differently. He had several sets of photos, one typically earlier and without and another later with make-up… More importantly, students were 23-24 at that point.
But all that became obsolete when most students ended up studying and finding jobs abroad and that effect dominated everything. That professor moved to study accents.
Interaction terms ftw :-)
I would postulate that every analysis becomes more accurate if you consider the interactions ... but at the same time it makes interpreting (and presenting) the results more complicated.
If women outnumber men in HR departments and at school, being pleasing to a woman's eye would be helpful in advancing through life more so than being pleasing to men.
If the effect is stronger for men than for women this could also support the more established notion that there are more social expectations on women to "look the part", i.e. wear makeup and "the right clothing" (which covers a much wider gamut than what is expected from men, which depending on white collar career path is either a casual suit or t-shirt and jeans/khakis). Attractiveness alone is not a guarantee for higher status and may in some cases even have the opposite effect by drawing unwanted sexual attention from male superiors or coworkers.
Same company, one of the smartest, friendliest and hardest working guys I knew was short (maybe 5'6" and completely bald) - he deserved to be a VP or higher imo, but never made it very far up the chain and left - also not a coincidence.
Yes men can have unrealistic standards at times but if you ever go to your local strip club and see the low attractiveness of your average striper you will notice men can be a lot more "accepting".