They actually mention it in the article that people tend to choose more reasonable numbers about groups they don't have much interaction with. Or when they do have a lot of interaction with one group, they push that experienced percentage down closer to the middle.
So like, if you're white and maybe 5% of people you interact with normally are black, you'd be more likely to assume that your experience is actually biased because you see more white people, so you pump the number up to try and get closer to the mean that you see as the "default".
There are definitely racist connotations in some cases, but a lot of it is just the psychology of straw polls like this.
The really telling thing is that this means people are guessing pretty wildly based on their experience (which includes propaganda/media consumption). All this survey really tells us that at best most Americans are really uninformed and at worst (which is the case) are incredibly misinformed.
They took the average of everyone's percentages and reported that. Averages of percentages are already a bit weird, but I think they didn't sanitize their data at all. So the 5% or so of survey participants who just answer the worst possible answer (100% of Americans are trans, etc) are pulling all the averages towards the middle.
This has to suggest an issue with the methodology right? Maybe people not understanding percentages?
deleted by creator
They actually mention it in the article that people tend to choose more reasonable numbers about groups they don't have much interaction with. Or when they do have a lot of interaction with one group, they push that experienced percentage down closer to the middle.
So like, if you're white and maybe 5% of people you interact with normally are black, you'd be more likely to assume that your experience is actually biased because you see more white people, so you pump the number up to try and get closer to the mean that you see as the "default".
There are definitely racist connotations in some cases, but a lot of it is just the psychology of straw polls like this.
The really telling thing is that this means people are guessing pretty wildly based on their experience (which includes propaganda/media consumption). All this survey really tells us that at best most Americans are really uninformed and at worst (which is the case) are incredibly misinformed.
They took the average of everyone's percentages and reported that. Averages of percentages are already a bit weird, but I think they didn't sanitize their data at all. So the 5% or so of survey participants who just answer the worst possible answer (100% of Americans are trans, etc) are pulling all the averages towards the middle.