FWIW, I'm sure you're aware of the OKCupid data from before the Great Awokening showing that, indeed, white men and Asian women were most rated most attractive by most races, and Asian men and black women least. The exceptions were that black men were rated as most attractive by black women.
The latest numbers I can find are from Quartz, in 2013, which find similar things:
Exception seems to be that now black men are at the bottom of the heap (except again for in the eyes of black women), and there is a preference of Asian men for Latina women; I don't know what this means or why it occurs (though there may be some delicious fusion restaurants in our future).
I am aware of this data. The issue with using it to conclude there are racial preferences is that the distributions of attractiveness might just differ by race above and beyond an impact of race itself. It's hard to tell, which is why more research needs to be done.
The problem is that attractiveness is a subjective, not an objective quality. There's enough correlation between people's individual ratings of people's attractiveness for us to say people are 'attractive' and modeling agencies and Hollywood to make money, but there really is heterogeneity among what people find attractive. Otherwise the same groups would come up on top *all* the time, and they don't--we see black women rating black men highest, after all.
There's also the whole other-aspects-than-beauty quality for women in particular; if they are judging white men as higher status (as would be quite consistent with what we know about female hypergamy) that would be a separate thing from pure attractiveness.
How you'd disentangle all that with something as un-PC as this where people are lying all the time, even to themselves, I have no clue.
Those photoshopped “unattractive” faces look like they have developmental abnormalities or recovered from facial reconstruction surgery after disfiguring injuries. Either of those possibilities has a much greater effect than just being “unattractive”.
I bet AI image generators would help here if used carefully. Could adjust the prompts to get just the right range of images.
The problem with AI generation is that they cannot, for the love of god, produce truly ugly people. They might create disfigured. Or create a monster. Or add stylish scars. But they cannot grasp the truly ugly. The mundane balding, bad skin, slight second chin. AI can draw grotesque, but it cannot draw ugly.
Interesting. The visual comparison between white and black gets even more complicated than it seems. I haven't read the attractiveness studies, but IIRC there was a study many years back examining police threat assessments of suspects that had a surprisingly deep analysis of factors. The one I found most compelling was the conclusion that suspect race didn't matter much as a direct factor. This contradicted early results showing that darker skinned suspects were correspondingly rated as more hostile and threatening than white peers. Later portions of the study included variations in lighting conditions as an experimental factor and a researcher noticed that although the apparent bias occurred when both were under the same lighting, it disappeared when relative visibility was accounted for. In other words, white suspects in dim lighting were rated just as threatening as black suspects in average lighting and both were rated relatively equally in very bright lighting (when both were clearly visible) or near darkeness (where both were barely visible). Threat ratings ultimately corresponded primarily to visual ambiguity, not race per se. This matched the larger body of research that shows that pretty much any factor that makes it harder to read faces clearly (such as clown makeup, masks, etc) tends to provoke higher perceptions of hostility. It's an unfortunate finding in the original context (darker skin really does unfairly make suspects seem more threatening in the average and dim lighting relatively common in many encounters with police), but it did somewhat amusingly provide a science-based endorsement that Blacks should consider wearing bright colors, white gloves, removing sunglasses, and otherwise seek to maximize how visible they are, especially their faces and hands.
Another study investigating police use of force found a similar effect: suspects in poor lighting were assessed as bigger, stronger, and tougher than the exact same person shown under better lighting. This built on medical studies that found that people, including doctors, would estimate black patients' pain thresholds as higher and injuries as less severe than equivalent white patients. Once again, visual ambiguity seems to have been the key factor: relatively less visible injuries and harder to read signs of pain resulted in reduced empathy and underestimation of the severity of wounds, corresponding to officers assuming that higher levels of force would be required to physically stop an individual.
All of which is to say that even just finding equivalently "attractive" white and blacks won't necessarily replicate, again, if relative lighting and perceptions of physical ability aren't taken into account. If the lighting is the same, darker skin may still read as more ambiguous and therefore potentially more hostile. Likewise, if the shared interest is a physical activity, the difference may lead to relative overestimations of the black person's physical capabilities. For example, if the shared interest is running or working out, there could be a perception that their pace would be too mismatched or their workouts too hardcore, which admittedly may also be a biological fact given the genetic differences that have resulted in Black athletes being disproportionately represented in sports involving running. For the purpose of attractiveness studies, it might be best for researchers to avoid physical activity contexts in framing the pictures and in profile interests.
Anyway, that's just my 2c. Sorry I don't have links for those studies, it's probably been over a decade since I read them and I don't recall the sources. I just figure that if chroma bias matters in both medical assessments and law enforcement assessment contexts, there's a good chance it's relevant in dating assessments too.
FWIW, I'm sure you're aware of the OKCupid data from before the Great Awokening showing that, indeed, white men and Asian women were most rated most attractive by most races, and Asian men and black women least. The exceptions were that black men were rated as most attractive by black women.
The latest numbers I can find are from Quartz, in 2013, which find similar things:
https://qz.com/149342/the-uncomfortable-racial-preferences-revealed-by-online-dating
Exception seems to be that now black men are at the bottom of the heap (except again for in the eyes of black women), and there is a preference of Asian men for Latina women; I don't know what this means or why it occurs (though there may be some delicious fusion restaurants in our future).
I am aware of this data. The issue with using it to conclude there are racial preferences is that the distributions of attractiveness might just differ by race above and beyond an impact of race itself. It's hard to tell, which is why more research needs to be done.
They might indeed.
The problem is that attractiveness is a subjective, not an objective quality. There's enough correlation between people's individual ratings of people's attractiveness for us to say people are 'attractive' and modeling agencies and Hollywood to make money, but there really is heterogeneity among what people find attractive. Otherwise the same groups would come up on top *all* the time, and they don't--we see black women rating black men highest, after all.
There's also the whole other-aspects-than-beauty quality for women in particular; if they are judging white men as higher status (as would be quite consistent with what we know about female hypergamy) that would be a separate thing from pure attractiveness.
How you'd disentangle all that with something as un-PC as this where people are lying all the time, even to themselves, I have no clue.
Those photoshopped “unattractive” faces look like they have developmental abnormalities or recovered from facial reconstruction surgery after disfiguring injuries. Either of those possibilities has a much greater effect than just being “unattractive”.
I bet AI image generators would help here if used carefully. Could adjust the prompts to get just the right range of images.
The problem with AI generation is that they cannot, for the love of god, produce truly ugly people. They might create disfigured. Or create a monster. Or add stylish scars. But they cannot grasp the truly ugly. The mundane balding, bad skin, slight second chin. AI can draw grotesque, but it cannot draw ugly.
How many times have you written out a post to near completion and the applet deleted it?
Once.
Interesting. The visual comparison between white and black gets even more complicated than it seems. I haven't read the attractiveness studies, but IIRC there was a study many years back examining police threat assessments of suspects that had a surprisingly deep analysis of factors. The one I found most compelling was the conclusion that suspect race didn't matter much as a direct factor. This contradicted early results showing that darker skinned suspects were correspondingly rated as more hostile and threatening than white peers. Later portions of the study included variations in lighting conditions as an experimental factor and a researcher noticed that although the apparent bias occurred when both were under the same lighting, it disappeared when relative visibility was accounted for. In other words, white suspects in dim lighting were rated just as threatening as black suspects in average lighting and both were rated relatively equally in very bright lighting (when both were clearly visible) or near darkeness (where both were barely visible). Threat ratings ultimately corresponded primarily to visual ambiguity, not race per se. This matched the larger body of research that shows that pretty much any factor that makes it harder to read faces clearly (such as clown makeup, masks, etc) tends to provoke higher perceptions of hostility. It's an unfortunate finding in the original context (darker skin really does unfairly make suspects seem more threatening in the average and dim lighting relatively common in many encounters with police), but it did somewhat amusingly provide a science-based endorsement that Blacks should consider wearing bright colors, white gloves, removing sunglasses, and otherwise seek to maximize how visible they are, especially their faces and hands.
Another study investigating police use of force found a similar effect: suspects in poor lighting were assessed as bigger, stronger, and tougher than the exact same person shown under better lighting. This built on medical studies that found that people, including doctors, would estimate black patients' pain thresholds as higher and injuries as less severe than equivalent white patients. Once again, visual ambiguity seems to have been the key factor: relatively less visible injuries and harder to read signs of pain resulted in reduced empathy and underestimation of the severity of wounds, corresponding to officers assuming that higher levels of force would be required to physically stop an individual.
All of which is to say that even just finding equivalently "attractive" white and blacks won't necessarily replicate, again, if relative lighting and perceptions of physical ability aren't taken into account. If the lighting is the same, darker skin may still read as more ambiguous and therefore potentially more hostile. Likewise, if the shared interest is a physical activity, the difference may lead to relative overestimations of the black person's physical capabilities. For example, if the shared interest is running or working out, there could be a perception that their pace would be too mismatched or their workouts too hardcore, which admittedly may also be a biological fact given the genetic differences that have resulted in Black athletes being disproportionately represented in sports involving running. For the purpose of attractiveness studies, it might be best for researchers to avoid physical activity contexts in framing the pictures and in profile interests.
Anyway, that's just my 2c. Sorry I don't have links for those studies, it's probably been over a decade since I read them and I don't recall the sources. I just figure that if chroma bias matters in both medical assessments and law enforcement assessment contexts, there's a good chance it's relevant in dating assessments too.
Super useful and interesting, thanks!