Conversation

Aella
@Aella_Girl
This is why you should always be skeptical of all research, ever
Quote Tweet
Spencer Greenberg 🔍
@SpencrGreenberg
Survey results can be used to glean genuine insights about how people think OR be used by biased or unscrupulous actors to mislead or manipulate the public. We ran an experiment to show how such distortion can work, with dramatic results! Our writeup: clearerthinking.org/post/our-resea
Image
12
303
cyn
@w0rldw1dew3b
Replying to
@Aella_Girl
working in a call center that focuses on doing telephone surveys, i can vouch that they are very hardcore about you reading everything verbatim regardless of whether it sounds right or not out loud. it's a huge thing for them and i totally get it
1
3
Orual
@NaamahOrual
Replying to
@Aella_Girl
Also, differences in the sample population! If you’re comparing male and female people in the Air Force, say, but a much larger percentage of the men than the women are on active duty vs reserves/National Guard, that is going to affect results for almost every metric.
Alex
@MephrosListens
Replying to
@Aella_Girl
It is considered incredibly important… Survey design is basically minimize cost and variance, and learn how to design the survey questions to not lead people on. If they aren’t taking the design portion seriously that’s terrible…
Image
3
Jameson Payne
@JamesonPayne5
Replying to
@Aella_Girl
Read "The Nature and Origins of Mass Opinion" by John Zaller. Eye-opening book on public opinion and surveying. What we call "public opinion" is a basically epistemically unknowable variable that is based on extremely short-term recall and, as this post shows, framing bias.
fahadalsaud.eth
@CryptoGrampa
Replying to
@Aella_Girl
Wow…so you mean asking 859 people a question, and extrapolating the responses to be indicative of a country with a population of 343,709,552 isn’t necessarily accurate? Woah. Mind blown.
Show additional replies, including those that may contain offensive content
Show