The Other Sexual Orientation (blog)

[filthytrancendence]

I read most of Aella's blog posts, and highly recommend perusing the blog at large. Her stuff around polyamory is particularly insightful, if you have any interest in hearing from an articulate person of that persuasion.

This one came into my email today, and I thought it might make for an interesting discussion here.

https://aella.substack.com/p/the-other-sexual-orientation

She's got a way of articulating and also demonstrating via data things I suspect but don't usually have language to articulate. This is an excellent example. I'm sure there's people who will bristle at the idea that there's more than one axis to view sexual orientation on, but I think there's a lot there worth exploring. Both in helping us to develop compelling characters and understanding niche audiences.

 

[TheRedLantern]

A lot of things made more sense after reading that.

 

[SimonDoom]

I enjoy reading her. She's obviously highly intelligent, and she's open-minded and willing to go where the data take her.

 

[TheLobster]

Careful, the last time I brought Aelia’s work on this forum I got an earful about how only peer reviewed research is trustworthy, and how this is just some biased internet survey.

Incidentally, that was also in the context of dom/sub being a much better descriptor of some people’s sexuality than homo/hetero. Some individuals don’t seem very keen on having their preconceived notions challenged, especially when it goes against the grain of “accepted wisdom” about privilege and discrimination.

 

[SimonDoom]

Careful, the last time I brought Aelia’s work on this forum I got an earful about how only peer reviewed research is trustworthy, and how this is just some biased internet survey.

Incidentally, that was also in the context of dom/sub being a much better descriptor of some people’s sexuality than homo/hetero. Some individuals don’t seem very keen on having their preconceived notions challenged, especially when it goes against the grain of “accepted wisdom” about privilege and discrimination.

I keep this in mind when I read her stuff. But I think she's a lot more intelligent and nuanced in dealing with data than many others with better credentials.

 

[AwkwardlySet]

It's an interesting read, and one that goes along with my own views. I see everything in our sexual nature as a spectrum. The homo/hetero one is the strongest one, in my opinion, but it's hardly the only one.
The dom/sub powerplay is also one. The consent/non-consent as well. Tender/rough. Romantic/physical only. And so on.

But more than all this data, I appreciate the idea to keep an open and curious mind rather than sticking to either traditional or very modern concepts of sexuality. Despite the very long time we've been at it, we haven't figured all of this out, not by a long shot.

 

[SinclairGroupLLP]

This was fascinating.

 

[TheRedLantern]

Careful, the last time I brought Aelia’s work on this forum I got an earful about how only peer reviewed research is trustworthy, and how this is just some biased internet survey.

TLDR
As far as I can tell, this was a good analysis done on good data. Like every analysis of every data set that's ever been done in history, it has a number of caveats. Dismissing this as invalid because of them is, in my STRONG opinion, an emotional and not a rational, statistical reaction.

The criticism is fair, although not valid, IMO.

It's fair because she shares her data and it's horribly biased. The overwhelming source for the data was "Other" aka TikTok where the average age was 22.6 years, compared to the US as a whole where the average age is 38.7 (median is 38.9, so pretty comparable). One thing that jumps out at me immediately is the "cisness", which ranges from 68% to 88% (she defines this as "I asked them their gender and their birth sex, and they are 'cis' if those two match"). Compare that to the overall US population which is closer to 97.4%.

I'm comparing to the US population and not the world as a whole because I think US data collection before Jan 2025 was the most thorough in the world.

As a data analyst, I would challenge claims that her data is representative of anything other than the population of people who were likely to be exposed to a survey that she put out. However, it's the best data available and it's a large sample (approx 400K survey results although it looks like she only used 80K for this article), so IMO it's worth studying for undeniable trends.

In her article about the data, she raises a similar version of this challenge but pointing out that the largest component of the data comes from TikTok, where the content shown is highly and effectively targeted. That may be a strength (she suspects the algorithmically targeted users are "people who like to take personality tests" and not "horny men") or a weakness.

Having said that, i don't think "it's not peer reviewed" is a valid criticism. She published her findings, and she published her data collection methodology. I looked to see if she had a github repo that might have some Jupyter notebooks or R scripts along with the survey definition and responses, or at least a Docker image that could regenerate those artifacts, but didn't see one. So I would call her research "somewhat reproducible." If I had the time and the tolerance for social media, I could probably replicate her study although it might take a few months and maybe more than one try.

Her conclusions are straightforward and shouldn't take much more than college level stats to derive. She doesn't share statistical power or p-values, but her R-squared analysis (the "Relative Strength of Factors") suggests that she almost certainly computed those because she explains what that section means in a layman's version of the statistical language, but any data analyst, data scientist, or statistician knows what she's talking about and knows that she is "one of us."

The final test, and this isn't scientific or mathematical, is "does the analysis make predictions that make sense?" To me, to does. A few results are very clear and don't need math to see. "Men are much more likely to LOVE blowjobs than women." "Women are much more likely to identify themselves as subs than men, women are much more likely to identify themselves as subs than Dommes." Those are, in effect, predictions that the data makes and I don't think anyone would mount a rational challenge against them when viewed as statistical statements. Like I said, this isn't scientific, but data analysts (at least ones who make it past their first year), almost always give results a "smell test" and this passes easily.

If I had to take this research to my manager, to a PhD data scientist, or to a business client, I would feel pretty confident in doing so.

 

[Bramblethorn]

TLDR
As far as I can tell, this was a good analysis done on good data. Like every analysis of every data set that's ever been done in history, it has a number of caveats. Dismissing this as invalid because of them is, in my STRONG opinion, an emotional and not a rational, statistical reaction.

The criticism is fair, although not valid, IMO.

It's fair because she shares her data and it's horribly biased. The overwhelming source for the data was "Other" aka TikTok where the average age was 22.6 years, compared to the US as a whole where the average age is 38.7 (median is 38.9, so pretty comparable). One thing that jumps out at me immediately is the "cisness", which ranges from 68% to 88% (she defines this as "I asked them their gender and their birth sex, and they are 'cis' if those two match"). Compare that to the overall US population which is closer to 97.4%.

I would generally concur with this. Unfortunately non-representative samples are an inevitable fact of life when investigating something like sexuality; generally the best one can do is to recognise the biases and figure out which kinds of conclusions are most likely to be affected by the data. As the saying goes, becoming a scientist because one wants certainty is like becoming a priest for the sex...

If she were using this survey to ask questions like "what percentage of men find blowjobs very/extremely erotic?" (looks like about 70% in her data), I would consider the results pretty much worthless. As already noted, the collection skews towards young people (whose sexual interests might not match older generations), and although this is harder to quantify, it almost certainly skews towards people who are relatively open-minded/curious about sexuality and perhaps to those who have a high sex drive. Those things are likely to correlate with stuff like "interest in blowjobs" and so the bias in the sample will probably have a big effect on "% who like blowjobs".

But her focus is more on bimodal patterns in the data and on how different variables correlate with one another. It's still possible for sampling bias to affect that kind of analysis but it's not as likely for it to create big effects.

For instance, if age has a positive correlation with some variable of interest like "interest in blowjobs", then anything that biases the sample towards younger people is also likely to bias the results towards lower values for that variable. But when we look at the relationship between those two variables in the data, we'd still be likely to see a positive correlation between the two. To conceal or reverse that correlation, we'd need some kind of second-order sampling bias which makes young blowjob-haters and/or old blowjob-lovers particularly unlikely to respond, while keeping young blowjob-lovers and old blowjob-haters willing to respond.

Having said that, i don't think "it's not peer reviewed" is a valid criticism. She published her findings, and she published her data collection methodology. I looked to see if she had a github repo that might have some Jupyter notebooks or R scripts along with the survey definition and responses, or at least a Docker image that could regenerate those artifacts, but didn't see one. So I would call her research "somewhat reproducible." If I had the time and the tolerance for social media, I could probably replicate her study although it might take a few months and maybe more than one try.

Making this fully reproducible would require publishing individual-level responses, which would risk exposing privacy of the respondents. While it doesn't collect direct identifiers, it looks like it collects enough to identify people at the level of e.g. "27-year-old trans male Pacific Islander now living in Luxembourg".

Peer review would be nice, but a lot of the people who'd be competent to give that kind of review would likely be unavailable or unwilling right now, given current goings-on in the US health and academic sectors and worldwide fallout of that.

Her conclusions are straightforward and shouldn't take much more than college level stats to derive. She doesn't share statistical power or p-values, but her R-squared analysis (the "Relative Strength of Factors") suggests that she almost certainly computed those because she explains what that section means in a layman's version of the statistical language, but any data analyst, data scientist, or statistician knows what she's talking about and knows that she is "one of us."

Also, the sample size is large enough that sampling (random) error is unlikely to be a big factor here; it'll be the systematic errors (bias etc.) that are a bigger concern.

 

[TheRedLantern]

But her focus is more on bimodal patterns in the data and on how different variables correlate with one another. It's still possible for sampling bias to affect that kind of analysis but it's not as likely for it to create big effects.

Good point. She really hits this. My suspicion is that you're right in that the sampling bias probably has only a minor impact on her conclusions that "there are these clusters of sexual interests that are distinct from sexual identity."

Now that I've thoroughly outed myself as an unrepentant math nerd, I find it extremely gratifying to learn that 40% of the men who responded said that they find "bimbos" to be "not at all sexually interesting."

 

[filthytrancendence]

I enjoy reading her. She's obviously highly intelligent, and she's open-minded and willing to go where the data take her.

At least we found something to agree on

As far as I can tell, this was a good analysis done on good data. Like every analysis of every data set that's ever been done in history, it has a number of caveats. Dismissing this as invalid because of them is, in my STRONG opinion, an emotional and not a rational, statistical reaction.

This seems fair to me. I'm coming to it from a less mathematical perspective, and while I think Aella would characterize herself as more of a data wonk than I, I've always read this sort of thing from her as doing her best to use the tools she has available to test her own experiences with sexuality against a broader base of experience. I also suspect the kind of questioning she's doing is also the kind of questioning a more traditionally credentialed data scientist would be unlikely to even try to test. And her capacity to formulate these sorts of questions is significantly different due to her background as a sex worker. And in my experience, sex workers sure have a much broader and more interesting view of sexuality than basically anyone who has not been a sex worker.

Her strength is in finding the more interesting questions to ask, I think. For me the most interesting graph in this blog is the Relative Strength Factors, and the tidbit about this being very similar to how personality tests work. That's looking behind the obvious answers and allowing a more interesting question to form. Namely, that perhaps gay/bisexual/straight is only one of several clusters that define trends in individual sexuality. And I think she'd be the first to agree that more academic research into the topic would be extremely welcome.

 

[filthytrancendence]

Careful, the last time I brought Aelia’s work on this forum I got an earful about how only peer reviewed research is trustworthy, and how this is just some biased internet survey.

Incidentally, that was also in the context of dom/sub being a much better descriptor of some people’s sexuality than homo/hetero. Some individuals don’t seem very keen on having their preconceived notions challenged, especially when it goes against the grain of “accepted wisdom” about privilege and discrimination.

Sure, I expect as much. There's a whole slew of people that will categorically dismiss everything she ever says because she was an escort. And a different slew of people that will do so because she's a woman who is not traditionally feminine and extremely confident and open about her sexuality.

It's just, I'm not that interested in what those sorts of people think

 

[TheRedLantern]

And her capacity to formulate these sorts of questions is significantly different due to her background as a sex worker

That's so interesting. I had no idea what her background was at all. This post was the first I'd heard of her, and this article (and the one she linked where she goes into a little more depth about the raw data) are the only two of hers that I read.

Based on just that, what I saw was an experienced statistician/data analyst. I would have suspected either an academic or a corporate background, probably in marketing analytics where analyzing survey data happens a lot. I figured it wasn't politics because she would have run the survey differently.

I would not have guessed escort. Sounds like her clientele was among that 40% of men who are turned off by bimbos.

I am pleasantly surprised

 

[iwatchus]

Based on just that, what I saw was an experienced statistician/data analyst. I would have suspected either an academic or a corporate background, probably in marketing analytics where analyzing survey data happens a lot. I figured it wasn't politics because she would have run the survey differently.

I would not have expected academic at all.To me, her stuff feels like a lot of self taught "Do my own research" types. You can get very good questions out of that community, but the analysis is often significantly flawed, glossing over issues that would argue against their point. I know many academics (in many fields) scream about the amateurs being in their field, as if only they were truly unbiased observers of the truth, which they are from from.

I actually give her more credence that she was an escort. While her own sample is obviously skewed, she presumably has a much larger collection of personal experiences than most of us, or almost any academic sexologist. If she was successful at her job, she understands what a broad range of her johns want. That is a good starting point.

I am still pondering this one. I may ask my son, who is a data scientist by training and has better statistical analysis chops than me. It's just an awkward discussion to have with your son, even your adult son. He is still recovering from learning I write here.

 

[Escierto]

One thing I noticed as I read the article was this statement: "...since males and females having fetishes for each other’s bodies results in reproduction, and thus has gotten real popular and become a wholesome cornerstone of civilization and media and identity."

This is the first time I have seen anyone refer to this as a fetish. Men like me who desire women with a penis are ALWAYS labeled as fetishists but men who desire women with a vagina are NEVER labeled as fetishists. As she says this fetish is a wholesome cornerstone of civilization whereas my fetish is just a perversion.

 

[filthytrancendence]

I know many academics (in many fields) scream about the amateurs being in their field, as if only they were truly unbiased observers of the truth, which they are from from.

Seems to me this is a pretty significant problem in the world. Far more so than any amateur who is open about the fact they're an amateur asking and attempting to answer the questions that interest them.

That's not a debate I'm trying to have here. I just think it's worth pointing out. She is and always will be marginalized from polite society. She will never be taken seriously by the sort of person that's making these complaints. I'd much prefer she ignores them and keeps asking interesting questions for us to ponder.

 

[Baztrachian]

Having said that, i don't think "it's not peer reviewed" is a valid criticism.

Agreed. If you view this as poll-based research then peer review is irrelevant as to the results of the poll.

It's like peer-reviewing an election. No matter how the 'experts' feel about the numbers the numbers will speak for themselves.

 

[Escierto]

The other thing I noted was where she said, "Roughly 80% of women in my data are sexually submissive, which you could say is default female sexuality." I read that and I was like, yes, this explains it.

I am a man and I have no desire to change my body but I am sexually submissive in a female way. Instead of a default male sexuality, I have a default female sexuality. (I guess this explains why I have never enjoyed receiving blow jobs but I love to eat pussy).

It explains why I am so happy to be the woman in bed for my trans girlfriend (who loves to receive blow jobs).

I was telling her today about how she used to come to bed drunk and pull my undies off while I was sleeping so she could have her way with me. She was embarrassed about it but I told her that I belonged to her and she could always use my hole to satisfy herself.

 

[filthytrancendence]

One thing I noticed as I read the article was this statement: "...since males and females having fetishes for each other’s bodies results in reproduction, and thus has gotten real popular and become a wholesome cornerstone of civilization and media and identity."

This is the first time I have seen anyone refer to this as a fetish. Men like me who desire women with a penis are ALWAYS labeled as fetishists but men who desire women with a vagina are NEVER labeled as fetishists. As she says this fetish is a wholesome cornerstone of civilization whereas my fetish is just a perversion.

Yeah. She's using fetish in a very neutral way there, and you are correct to observe there is a societal trend to using it as an insult. It's a subtle but dramatic shift in viewing sexuality. Fetish is just whatever arouses you. There's a lot of motivated reasoning going on at the societal level about who is labeled perverse.

I think we can all agree perversion is a useful concept. When fetish is attached to people or things that cause harm, that's a problem. But you have to actually demonstrate harm for that to be a valid label, in my view. Sometimes the harm is obvious. Rape. Pedophilia.

But most of the things labeled perverse are only potentially harmful. Just like literally every possible kind of sexual activity. And a lot of the things labeled perverse, like the preferences you're sharing, are only labeled as such because they are different from what is widely accepted as normal. And that's just horseshit.

 

[iwatchus]

Seems to me this is a pretty significant problem in the world. Far more so than any amateur who is open about the fact they're an amateur asking and attempting to answer the questions that interest them.

I am at a "Pox on both sides" level on this one. And agree that I have no desire to take that further. I was just trying to point that there are two camps and they both have issues. But, when push comes to shove, I will usually stand with the academics' side.