Alright you know the drill - I built a giant survey that went viral, resulting in my current dataset of 640k respondents. Demographics skew towards being mostly young liberal western women, and align pretty well with “the kinds of people who take personality surveys online.” The dataset isn’t perfect, but seems relatively good anyway - for example, I keep accidentally finding stuff in my survey that replicates existing research, which means if there’s skew in the survey it’s likely to be the same type of skew that we’d fine in normal academic publications.
In my survey, I asked people “Compared to other people of your same gender and age range, you are [Significantly less attractive — Significantly more attractive]”
So let’s look at the women who report being hot! There’s 268,000 of them for this sample, and I’m looking at women between the ages of 19-29.
First off - how accurate is women’s assessment of their own hotness? We don’t know for sure. Hotter women do report weighing less:
So there’s a signal there of something at least!
(-3, the leftmost bar, is the least-attractive-reporting people; the 3 on the right is the most attractive)
Still, we might also be getting a bit of noise from things we don’t intend to measure, like lies to herself about how hot she is or is hotter than she thinks but is in denial about it. Keeping that in mind, let’s forge ahead!
Mid women have the least PMS
This one’s pretty weird. PMS is recorded where 0 is no pms, 1 is ‘slight’, 2 is ‘moderate’, 3 is ‘severe’. In general there’s barely any variation, but the little bit that’s there is bimodal. What?? My sample is big enough that this is almost certainly not noise.
I’m gonna zoom in the y axis for more graphs to look at the differences. This is a deliberate choice, keep an eye on the y axis yourself, don’t yell at me.
Hotter Women Are More Cisgender, Mostly
This is interesting - in general, the hotter a female reports being, the less likely she (he?) is to report being trans. There’s a bump at the most attractive category. My first thought is - are we catching a troll type response, where people pick the weirdest category out of each question, so you see strange correlation behavior at the extremes?
But I don’t think so - I already cleaned my data pretty hard of a lot of other troll responses (like, high variance responses among similar questions, or eliminating the tail end of some other easy-to-troll questions), so I’d be a bit surprised if this were entirely a troll phenomenon.
My current top guess is “most transmen view themselves as unattractive compared to others, but there’s a small cohort that view themselves as maximally attractive compared to others.” I suspect this is likely a ‘comorbid with mental illness’ thing, but I don’t know for sure. Maybe there’s legit a chad cohort out there absolutely slaying.
Hotter women grew up in higher class families
I asked people “Growing up, your family was [under—upper] class, compared to most of the people in your country”.
The spectrum was 0 (lowest) to 6 (highest), so you can see most answers here fall between 2-3 (or lower-middle to middle class).
This is nearly linear; the women reporting growing up in higher class households, also reported being hotter than most people. Seems like some evidence for ‘You aren’t ugly, you’re just poor!’
Mid women were spanked the least
…hahaha what? …. What though?
This was a question asking ‘how frequently were you spanked as a kid (pre 14) as a form of discipline?’ And the answers were a 0 (never) to 4 (very regularly) range.
Obviously, the y-axis is zoomed here and the absolute difference isn’t huge (just a 0.33 gap out of a 5-point spread), but it’s still not happening by accident here.
I have no idea how to interpret this. Spanking is correlated with class, like lower class families spank more, so maybe this is an artifact of that, but… why are the hottest women reporting an uptick?
If someone reported that they were spanked, then I gave them a question asking about how painful or severe the spankings were. The chart for this is even more clear:
Super weird. Often if you get strange patterns it’s due to an obvious confounder, like “oh it’s just that older people are both weaker and also more tired, so when we see that weak and tired correlate it might just be age.” But when something’s bimodal I find it less likely to be some secret hidden third thing in the data.
Still, the total difference is pretty small!
The jews are sexy
I asked people what religion they predominantly grew up in. Of these, the areligious are either the most uggos or have the lowest self esteem (or both). Jews are either the hottest or have the highest self esteem (or both).
Interestingly the Muslims are also doing pretty well.
The absolute severity of religious upbringing didn’t seem to have any cool-enough impact on hotness to show here, which is interesting.
As a proxy for traditionalness of household, I asked about how bad gender role violations were when they were growing up. Like, were people chill with nontraditional households, or would they get mad if a woman worked outside the house?
0 was ‘so normal nobody noticed, and 5 was ‘community would evict you.’ Again, the absolute difference isn’t huge, the biggest gap being 0.18/5.00 but the small trend is mildly interesting.
Sometimes people are like ‘aella why are you sharing results about something that’s barely a difference at all’, but I sort of feel like everything can be a clue that informs your theories about the way the world works. If you have a thousand tiny clues, it might eventually build up to good evidence about something.
Mid women had the most loving families
What does this mean?
This is obviously connected to the spanking chart; in general, people who report being spanked also report being physically abused.
All my theories for this are wild. Is it that of people who spank/abuse a lot, they tend to be either low class (casual slapping around alcoholic mother vibes) or high class (rigid proper family)?
Or is it that people rating themselves as very attractive tend to be more sensitive to how they’re perceived, and this also somehow correlates with a greater tendency to overreport bad things that happen to them? In my data, women report generally worse experiences than lower class than men.
My guess is this is due to neuroticism, and there’s at least a consistent correlation here - people who report being less relaxed, also report being spanked more.
So, are hotter women more neurotic? If the most neurotic women put their neuroticism points into either being hot or convincing themselves they’re hot, then maybe they also are more likely to report being abused?
Chill Women Are Hotter, Actually
…nope. This chart is the avg score of people’s answers to my two neuroticism questions, “I worry about things” and “I am relaxed most of the time (reversed)”.
This contradicts the hypothesis of neuroticism causing an uptick, and makes me think it’s less likely that the bimodal abuse graph is a reporting artifact.
But anyway - huh! The chiller a woman is, the more likely she is to view herself as hot. Cool.
Hotter women feel more powerful
This is the avg score of the questions "I don't have very much power over those around me", “I deserve more respect than I get”, and "If life is a game, then I'm losing".
Anxious Attachment Styles Are Least Hot
Who knows which way the causality goes, but women reporting an anxious-attachment relationship style also had the lowest hotness score.
Sociopaths report being the hottest, social anxiety the ugliest
The avg attractiveness self-rating overall was -.24, by the way!
Sociopaths report being the hottest. I’m super curious if this is true, or if it’s an artifact of sociopaths having artificially inflated self assessments.
Anorexics report being relatively hot; body dysmorphics and social anxiety people report being the least hot.
I’m kinda surprised that autism and schizophrenia are ranked so low. They feel relatively neutral to me, like less socially mitigated, so I would have expected them to be a lil higher.
High openness? Hot.
This is the openness score, which is the avg of two openness questions - "I have excellent ideas" and "I have difficulty understanding abstract ideas (reversed)”.
The total spread is -6 to 6 (as it is with all variable avgs) so the biggest gap is 2.48/6.00. This is pretty huge! What?
They’re… smarter?
In my survey, I asked people “have you ever had an officially administered IQ test?” If they said yes, then I gave them a question asking what the score they received was.
This isn’t a perfect measurement, it’s still a self report, and you’ll notice the average reported score is quite high. I do think the avg high score isn’t that damning though; I suspect that most people who were given IQ tests were given them because they seemed really smart, and that people who know their score are disproportionately smarter than people who don’t.
I’d separately be pretty surprised if none of this trend was from people being insecure/arrogant about themselves in general, though, where if someone feels bad about themselves they will be less likely to inflate an IQ score and more likely to pick 'less hot than others’.
But still - I think openness is correlated with IQ, right? And openness questions are less ego-attached than IQ questions are, but those show an even stronger trend.
Mid women are… the most agreeable?
Oh man this is fascinating.
This is the avg of "I feel little concern for others (reversed)" and "I sympathize with others' feelings"
The u-shape is so cool. It does feel a little sociopathy-adjacent. If you think hot women are bitches, then this graph suggests you should also think of ugly women as near-equal bitches. Maybe both ugly and hot women get ignored or pursued, and both of the ends of that spectrum cause hostility?
Hotter, more conscientious and extroverted
The conscientious questions were "I shirk my duties (reversed)" and "I like order". Extroversion questions were "I am quiet around strangers (reversed)" and "I am the life of the party"
It’s a bit intuitive that hotter-reporting women are more extroverted; they get more socially rewarded! It’s less intuitive that hotter women are a bit more conscientious. maybe an artifact of less mental illness?
Well, they kinda are less mentally ill
???
The hotter the woman, the fewer mental illnesses she selected from the list of mental illnesses… kinda. It’s u-shaped here, which is confusing. Maybe the max hotness option is a weird artifact somehow? Maybe the women who pick being max hot are just straight up insane, but if you’re mid-to-cute you’re not insane? I don’t know. I’m not sure how this makes sense with the rest of the data and I’m curious if you guys have any theories.
Mid kids don’t get molested?
I asked if people experienced sexual assault in childhood, and people answered on a scale from no (0) to yes, severe (3).
This is prob part of whatever phenomenon is causing the bimodal distribution in spanking and abuse reports.
I’d like to point out that it’s really weird that we’re not seeing an uptick of neuroticism among the hottest women here. If the hottest women have more mental illnesses, get abused in childhood more, it seems likely it’s caused by something that also causes neuroticism. But mental illness without neuroticism is bizarre.
Hot women do get assaulted more tho
This one is cool - the hotter someone reported being, the more likely they were to report being sexually assaulted. This was a 0 (not assaulted) to 3 (severely assaulted) scale. This does seem to imply that attractiveness does influence the likelihood of men assaulting you!
Also, belief in the supernatural
This is “I find the existence of the supernatural to be plausible, e.g., ghosts, multi-dimensional beings, energy healing, astrology, etc.”
This isn’t centrally astrology, but is probably a good proxy for girls who are into astrology. The ‘hot girls are into astrology’ then looks like it is slightly true - but it’s more true that unattractive girls are into astrology.
No idea if this is tied to the mysterious other bimodal trends we’ve been seeing, though.
Liberalism?
Social and economic liberal/conservatism questions are basically the exact same graph, a weird M shape. Keep in mind absolute gap is still relatively small - about 0.2/5.00. Maybe we’re seeing the middle depression as “people who are noncommittal/average about questions in general?”
But still - both the most unattractive and the most attractive categories report the greatest conservatism. I’m so confused.
—
In the upcoming pt 2 post, we’ll look at how women’s hotness corresponds to relationships and sexual dynamics. Do they break up with people more? How poly are they? Are they more committed? What’s their bodycount?
Subscribing is free, and you’ll get the upcoming posts straight to your inbox.
I think you should've asked them to take 360 photos of themselves and hire a guy in islamabad to rate all of them
The recurring U-shape might be a response style effect - people who rate themselves as 'average' in looks might just be people who generally see themselves as 'normal' across the board. You mention this possibility in the politics questions but it could hold for others like agreeableness, mental illness, how loving or traditional your family is... This wouldn't explain every pattern, though! I agree that small effect sizes can be very interesting especially in such large datasets.