Conversation

so this did the rounds today, i suspected selection bias in aella's samples bc the claims made seem a lot different than other data, and i took a look at her article, the fact that men and women in her sample watched comparable levels of porn, says it all. huge selection bias

Quote Tweet

eigenrobot

@eigenrobot

Jan 27

Replying to @radfemswiftie

i dont know if your aware but rape fantasies are mostly a female phenomenon. see data from twitter.com/Aella_Girl/sta

Replying to

it wasn't that similar. It was more similar if you included erotica.

Replying to

i think it' fun that you gather data and stuff but it just triggers my autism bc of social science methodology and stuff, im not trying to throw shade at you. i think you're an interesting person even if I don't agree with your lifestyle

Replying to

also, my sample is 480,000 young people on tiktok; not totally representative, but likely more representative than most other studies. If our stuff contradicts good chance on priors they have more selection bias issues than I do

Replying to

how do you figure? its an internet survey with no ability to verify the integrity of the data, which derives the data from a particular segment of the population within a narrow demographic window of specific userbases, narrowed by algorithm, and then further by interest

Aella

@Aella_Girl

Replying to

@dancerinspace

u should look into how most peer reviewed stuff in published journals get their survey data, if u think mine is bad those will give u a heart attack Also there are ways to help verify data integrity! I design my surveys specifically with that in mind.

6:15 PM · Jan 29, 2023

269

Views

Replying to

of course, i totally agree! a lot of survey data is heavily skewed and entirely useless, though there are still some gems out there. its still really cool that you're collecting data on this scale, and i definitely appreciate the demographic information you collected with it