Conversation

Twitter polls are no better than noise, they said. Concluding anything based off twitter polls is laughable and meaningless, they said. Real research is the kind that gets published in academic journals, they said. (look at that stable 2.2%!)

238

Aella

@Aella_Girl

Aug 15, 2022

twitter polls obvs aren't perfect, but there's lots of ways to get a good grasp on their limitations and draw pretty good conclusions from their data. I think people often dismiss them because they don't have enough "science aesthetic".

161

Replying to

and horrific selection bias lmao

Replying to

ah yes, the horrific selection bias that eliminates the meaningfulness of the slight increase of covid rates in a population over time

Replying to

oh for sure; that was just responding to the bit about people dismissing twitter polls in general. the slight increase is interesting - i mean there's not likely to be a whole lot of noise in the data. it'd be cool to see how well this lines up theoretically with actual rates

Aella

@Aella_Girl

Replying to

@grumpysavior

when i ask what gender ppl are, i don't think that the 85% male respondents means that 85% of all people are male. horrific sampling bias is only horrific if you're using it to make claims in the wrong contexts, which i am very careful to try not to do

5:27 AM · Aug 15, 2022

Replying to

oh i don't think you do - you just put a poll up there. but then everyone reading the poll interprets it. so if you have a poll that's male+A/male+B/female+A/female+B, people are gonna just draw the simple conclusion "looks like women are X% more/less likely to be B than men"