A RANT: Are my polls unscientific? Depends on what you mean - but largely I hear the word 'unscientific' used in place of 'does not have enough sciencey aesthetic' - for example often people say my twitter polls aren't 'scientific' by pointing to some of my joke polls 1/
Conversation
Replying to
Joke polls aren't a 'science aesthetic' (and neither is twitter in general), so it's basically people pointing at my twitter polls and going "Look, there's no lab coats, this is untrustworthy."
But this is a TERRIBLE way to think about science!! 2/
1
1
67
You don't need a lab coat or a peer reviewed paper to do science. Science is just a method of figuring things out, of discovering bits of evidence that points in various directions, and YES twitter polls *are bits of evidence.*
4
3
74
They're not perfect evidence, ofc - most of the people who answer my polls are white males 25-35 years old, so we can't call it a random sample - but tons of "science with a lab coat" gets a pass despite having worse sampling biases (and way lower sample sizes) than I do!
2
82
I *regularly* spot check my polls for accuracy. Sometimes I do two sets of polls - with just question A (e.g., politically right or left?) and then again adding in question B (e.g. do you value freedom or safety?) to see how much A answers change due to priming.
1
34
I also do series of polls to test the troll ratio, and have found troll answers increase significantly the less the answer has to do with personal identity; e.g., a series of increasingly complex math problems has *more* correct answers the more complex it gets.
4
47
Does this address all flaws? No! But evidence doesn't have to address all potential flaws, it just has to be evaluated in context. Plenty of research is done on way less reliable samples and nobody gives a shit cause it's done by credentialed people writing in the passive voice
4
2
58
I personally use my polls as light bits of evidence pointing in directions that might be useful to research more, kinda pilot-study-esque. Like, this poll is *so cool*, it's fascinating evidence pointing heavily in a direction we'd be amiss to ignore!
Quote Tweet
Do you have chronic anxiety? || Do you have stomach problems (like IBS, constipation/diarrhea, etc.)?
Show this poll
4
1
57
The people who call my polls unscientific have a really limited idea of what science. It's all around us, and it can be informal or unassuming or outright goofy, but as long as you're *finding bits of evidence and not overinterpreting* then congrats, you're doing science!
3
1
76
This 'science aesthetic is science' flaw is one of the things that bothers me the most about the way we culturally approach learning, and I'm planning on stabbing it in the face in a lot more detail in the book I'm attempting to write.
14
1
95
Replying to
hope to see you similarly throwing out all the published, peer-reviewed papers that don't have random sampling!
2
48
Show replies