One thing that might be worth breaking out is sex differences vs sexual orientation differences. For instance, feet are a male-skewed interest, but at least in my data when I broke it out by sexual orientation, straight men and lesbian women scored higher than gay men and straight women respectively. So feet might be a gynephilic interest, more than a male interest per se.
This is super interesting. Two data points surprised me in their lack of popularity and their gender composition, to the point that I wonder if they might get different results if phrased differently - specifically "older people" and "very overweight people." My sense is that you might get a lot more women for "older men" or "daddy", and a lot more men for "BBW".
One consideration is whether the strong negative association between tabooness and popularity could be _partly_ driven by people being less willing to express interest in fetishes to the extent that they believe they are taboo.
It could be interesting to include some things which might help test/account for this in future iterations (rot13'd in case you don't want people to know what the measures are for):
- Vapyhqr fbpvny qrfvenovyvgl zrnfherf
- Onlrfvna Gehgu Frehz (tbbtyrnoyr) / nfx guveq-crefbany dhrfgvbaf (ubj znal crbcyr qb lbh guvax ner vagrerfgrq va k)
Guvf znl erdhver fbzr vaqvivqhny yriry qngn gubhtu (v.r. nfxvat vaqvivqhnyf nobhg obgu nobhg vagrerfg va k naq gnobbarff bs k engure guna whfg ybbxvat ng nttertngr pbeeryngvbaf.
Could you say more about how you "did factor analysis... to find out the most predictive items to include"? The way I would have imagined a factor analysis working would to looking at associations within the set of taboo items, to see how they load onto different factors and how those factors relate to each other. But then I'm not sure why that would be finding the "most predictive" items to include.
Where are your magenta and red squares? It seems like a key point to your graph that women’s responses were more varied so as to not create any magenta or red squares. Whereas, for cis men, the responses were more homogeneous. What are your thoughts about this?
Fetish Tabooness and Popularity (v3)
Did you mean to say r = -0.75 (negative correlation)?
One thing that might be worth breaking out is sex differences vs sexual orientation differences. For instance, feet are a male-skewed interest, but at least in my data when I broke it out by sexual orientation, straight men and lesbian women scored higher than gay men and straight women respectively. So feet might be a gynephilic interest, more than a male interest per se.
This is incredible. Some of your best work. There's so many themes and groupings that can be seen here, and very easily too. Nice job!
PS it would be great to have an interactive version of this where one can filter the results based on various criteria! Maybe v4.0? 👉🏻👈🏻
This is a fantastic chart. So many little stories! Thanks for sticking it out and really polishing this.
This is super interesting. Two data points surprised me in their lack of popularity and their gender composition, to the point that I wonder if they might get different results if phrased differently - specifically "older people" and "very overweight people." My sense is that you might get a lot more women for "older men" or "daddy", and a lot more men for "BBW".
This is fascinating. I had no idea some of these things were kinks. Executions? How interesting.
Is the chart trying to tell us that romance and blowjobs are besties? ;-)
It also proves that Doms are the limiting factor for BDSM.
The male/female interest ratio for masturbation is inverted from what I would expect.
Zombies?
Interesting that the top right corner is empty. Basically there is nothing which has a high taboo rating and high interest rating.
One consideration is whether the strong negative association between tabooness and popularity could be _partly_ driven by people being less willing to express interest in fetishes to the extent that they believe they are taboo.
It could be interesting to include some things which might help test/account for this in future iterations (rot13'd in case you don't want people to know what the measures are for):
- Vapyhqr fbpvny qrfvenovyvgl zrnfherf
- Onlrfvna Gehgu Frehz (tbbtyrnoyr) / nfx guveq-crefbany dhrfgvbaf (ubj znal crbcyr qb lbh guvax ner vagrerfgrq va k)
Guvf znl erdhver fbzr vaqvivqhny yriry qngn gubhtu (v.r. nfxvat vaqvivqhnyf nobhg obgu nobhg vagrerfg va k naq gnobbarff bs k engure guna whfg ybbxvat ng nttertngr pbeeryngvbaf.
analisussy
tfw your kink is high-taboo, low-interest :\
Thanks for doing this!
Could you say more about how you "did factor analysis... to find out the most predictive items to include"? The way I would have imagined a factor analysis working would to looking at associations within the set of taboo items, to see how they load onto different factors and how those factors relate to each other. But then I'm not sure why that would be finding the "most predictive" items to include.
Love the multiple dimensions on the graph. Beautiful work
how ware can i read your list of taboos.mark
how ware can i read your list of taboos.mark
Where are your magenta and red squares? It seems like a key point to your graph that women’s responses were more varied so as to not create any magenta or red squares. Whereas, for cis men, the responses were more homogeneous. What are your thoughts about this?