Conversation

i have a list of animals ranked by the % of people who find them sexually attractive. i feel so powerful

356

Sep 3, 2022

Replying to

You do know twitter polls are a terrible source for accurate data? and your data is more likely a reflection of you than it is the general population? If not then this must be either some next level cope or a long running engagement bit.

Replying to

n=33k so far

My kink survey design methodology

This is a very detailed, very simple step-by-step walkthrough of my process for creating my currently-running fetish survey. It’s meant to be easily accessible/understandable to people with little...

Replying to

Just finished reading, I am speaking specifically about limitation #3. Your data will always be non random as long as you have online influence, this is not something you can compensate for. It will always be an aella survey even if you think data is coming from random places.

Replying to

and

You are and will always be subject to algorithmic sorting that only gets worse as you grow. Even shared polls and moving to adjacent groups on social media would be non random data.

Aella

@Aella_Girl

Replying to

@DvnnyyPhantom

only hoping you also level this same criticism at like 98% of all published social research!

7:14 AM · Sep 3, 2022

Replying to

I do! I also think you’re passionate about data which is cool! You just will eventually hit a wall where you realize you’re dealing with non random samples and the data can not be applied. So it may be best to focus more on sampling.

This Tweet was deleted by the Tweet author. Learn more

Replying to

and

Yes exactly, this is why I called the results more reflective of self than of gen pop. The argument that respondents are not coming directly from one’s following would still have to take into consideration the noise created by algorithmic content suggestion via social media.