Who Took My Big Kink Survey? Updated

Breaking down answers by survey source might confirm some stereotypes

I initially wrote up the summary here, when I had closer to 40k responses. Since then the survey went viral, mostly on tiktok, and I now have over 400k responses and still growing. Here’s the updated demographics!

I asked people “how did you find this survey”; I didn’t include tiktok as an option, so most of these went into ‘other.’ (I confirmed this by watching the ‘other’ category specifically balloon as tiktoks featuring my survey were gaining tons of views.

1. Other: 318,132

2. Twitter: 31,054

3. Reddit: 24,658

4. Discord: 16,312

5. Facebook: 3,741

6. Tumblr: 2,719

7. Fetlife: 2,386

8. Telegram: 2,070

For a TLDR summary: Compared to the world, respondents skew younger, more female, more liberal, middle class, queer, white (tho less white for the US population), western and heavily US based.

How do they differ from general population?

Picking a few lifestyle/demographic questions to look at! I first averaged male and female responses separately, and then took the average of those two numbers so that the data wouldn’t be skewed by different gender ratios.

Cisness

I asked people what their gender was and what their birth sex was; if they matched, they counted as ‘cis.’

Cisness of sources:

1. Reddit: 88%

2. Other: 85%

3. Fetlife: 85%

4. Telegram: 82%

5. Facebook: 81%

6. Twitter: 80%

7. Tumblr: 69%

8. Discord: 68%

Sex of sources:

What was the male/female breakdown from each source? Here’s the % of each source that reported being a biological female:

1. Tumblr: 91.5

2. Other: 78.8

3. Facebook: 61.1

4. Telegram: 52.7

5. Twitter: 51.8

6. Discord: 47.0

7. Fetlife: 46.1

8. Reddit: 43.4

Politics

Everyone was more liberal; I asked people to rate how economically and socially liberal they were on a -3 (conservative) to 3 (liberal) spectrum. I was really surprised to find “other” (tiktok) was the least liberal!

1. Tumblr: 1.84

2. Twitter: 1.59

3. Reddit: 1.57

4. Discord: 1.54

5. Fetlife: 1.45

6. Telegram: 1.40

7. Facebook: 1.30

8. Other: 1.29

Social class

I asked people what social class they were raised, on a spectrum from underclass/very poor (0) to elite class/very rich (6). All the groups averages, averaged together, was 2.5, split between lower-middle class (2) and middle class (3).

1. Twitter: 2.70

2. Telegram: 2.68

3. Reddit: 2.66

4. Other: 2.58

5. Discord: 2.57

6. Tumblr: 2.53

7. Fetlife: 2.49

8. Facebook: 2.42

Big5

This one’s a bit janky, so take with a grain of salt. I took 10 questions from the official big-5 test, for a total of 2 questions per dimension, to try to get a picture of people’s big 5 scores. This was only two questions, so it’s probably quite a bit noisier. This is probably why extroversion is weirdly low; I likely happened to pick two questions people were more likely to disagree with.

I also made my own powerlessness scale, with a score from 3 questions trying to get at how powerless people feel in life, asking people to agree/disagree with "If life is a game, then I'm losing", "I don't have very much power over those around me", and “I deserve more respect than I get”. The higher the powerlessness score, the more people agreed with those three statements. (I didn’t test or validate this scale at all)

Polyamory

I asked people their preferred relationship style, from monogamous (-3) to polyamorous (3).

1. Fetlife: -0.21

2. Telegram: -1.34

3. Facebook: -1.40

4. Discord: -1.43

5. Tumblr: -1.44

6. Twitter: -1.68

7. Other: -1.92

8. Reddit: -1.96

Age

People were young! Average age per source:

1. Other: 22.59

2. Discord: 22.98

3. Tumblr: 23.80

4. Reddit: 24.26

5. Telegram: 24.36

6. Twitter: 25.54

7. Facebook: 27.38

8. Fetlife: 31.78

BMI

BMI per source, not adjusting for age, which probably accounts for a lot of it:

1. Telegram: 25.47

2. Twitter: 26.00

3. Other: 26.60

4. Reddit: 26.70

5. Tumblr: 27.03

6. Discord: 27.12

7. Facebook: 28.08

8. Fetlife: 30.09

Childhood abuse

I asked if people were abused as a child, with binary answers of ‘yes’ and ‘no’ (more detailed questions about types and degree of abuse came later, if people answered yes). The exact question:

Were you (at least semiregularly) abused as a child? "Abuse" means cruel treatment, such as significant neglect, insulting, or causing physical damage. This question does *not* include spanking

% of people who responded yes:

1. Tumblr: 39%

2. Fetlife: 38%

3. Discord: 37%

4. Facebook: 36%

5. Telegram: 34%

6. Other: 31%

7. Twitter: 30%

8. Reddit: 29%

Location

Mostly western countries! The 5 most common sources:

1. US: 48%

2. Europe (non-UK): 17%

3. Canada, Australia, UK: 15%

4. South America: 5%

5. Asia: 2.3%

Ethnicity

I asked about ethnicity, with really basic options copied from a standard demographic survey form online.

Whiteness:

1. Fetlife: 73%

2. Reddit: 71%

3. Facebook: 70%

4. Telegram: 68%

5. Discord: 66%

6. Other: 65%

7. Tumblr: 65%

8. Twitter: 64%

Mental Illnesses

I gave people a list of mental illnesses and asked them to check whichever ones applied to them. This is the average number of illnesses people selected, per source

1. Tumblr: 3.15

2. Discord: 2.87

3. Facebook: 2.74

4. Fetlife: 2.68

5. Twitter: 2.38

6. Other: 2.38

7. Reddit: 2.36

8. Telegram: 2.03

Hotness

I asked people “Compared to other people of your same gender and age range, you are” with a scale from significantly less attractive (-3) to significantly more attractive (3). The average score per source:

1. Telegram: -0.03

2. Facebook: -0.04

3. Twitter: -0.04

4. Fetlife: -0.05

5. Reddit: -0.08

6. Tumblr: -0.13

7. Other: -0.16

8. Discord: -0.29

Some fetish spot-checking

How much does sexual interest in things vary per source? All of the following are avg ratings on a 0 (no interest) to 5 (extreme interest) scale. I picked four at random, before looking at the data in them:

Bestiality

1. Fetlife: 0.67

2. Telegram: 0.49

3. Discord: 0.48

4. Facebook: 0.38

5. Tumbr: 0.36

6. Twitter: 0.36

7. Reddit: 0.30

8. Other: 0.26

Incest

1. Fetlife: 0.94

2. Twitter: 0.72

3. Telegram: 0.71

4. Facebook: 1.48

5. Discord: 0.63

6. Reddit: 0.36

7. Tumblr: 0.58

8. Other: 0.27

Sensory

1. Fetlife: 1.75

2. Discord: 1.03

3. Tumblr: 0.92

4. Facebook: 0.88

5. Telegram: 0.85

6. Twitter: 0.73

7. Other: 0.72

8. Reddit: 0.71

Power dynamics

1. Fetlife: 3.39

2. Tumblr: 2.81

3. Discord: 2.63

4. Twitter: 2.44

5. Other: 2.41

6. Reddit: 2.36

7. Facebook: 2.29

8. Telegram: 2.16

At a glance, it looks like the “other” category - the biggest one - is not very kinky, but also roughly comparable to my reddit and twitter responders, which were my biggest sources before the survey went viral.

How Representative Is Tiktok, If It Uses A Targeted Algorithm?

It does use a targeted algorithm, and I’m pretty sure it targeted “people interested in personality surveys.” It didn’t target “horny men”, for example - my own tiktok specifically targets horny men for business purposes, but the sample that took my survey from tiktok were majority young and female. I have a friend who runs a very large personality testing website, and he also reported to me that most of the people who take his surveys are young and female. In general, young women are drawn to internet surveys that tell them something about themselves, and it seems very likely, based on priors and the data in my survey, that this is what the algorithm was targeting.

---

In general, I think the representativeness isn’t bad, as far as most surveys go. My main concern is westernness; I’d anticipate if I ran this same survey with primarily responses from India, for example, I’d see significantly different results.

I’m also a bit concerned about the young age. I can still easily see how age influences sexuality in my data (still tons of data from older people!), but if age does have a big impact on sexuality, I’d expect this to artificially skew any reported base rates. If this is the case I can probably control for age, I’m guessing?

I’m pretty happy to see that the demographic my survey went viral in - tiktok - seems a bit more representative of the US population than the demographics I had before. The biggest flag for this seems to be politics - trending more conservative, which feels like I’m somehow hitting a more general heartbeat of the population as opposed to the more liberal-specific online circles I’m in.