Americans overestimate minority population, underestimate size of majority groups

[DrScholl]

https://twitter.com/Travon/status/1633932471300673536?t=zrGQFDYLbPeVdJP5vqGsYQ&a...

This explains a lot.

[Хahar 🤔]

Normally when I see a chart like this it's about Europeans estimating the percentage of Muslims in their countries, which might make a point about xenophobia, but the fact that this pattern shows for every characteristic regardless of its political valence shows that this in fact doesn't explain anything except that people have trouble intuitively understanding population percentages.

[Sprouts Farmers Market ✘]

All of them are like 50-50 - you either are or you aren't.

Unless, you are like one-third Native American, but you need to be a math wizard to multiply those fractions.

[40% for military veteran might be the wildest thing I've seen. Even for draft era men, they ought to exclude women - and then 40% is still the average!]

[Diabolical Materialism]

Love the world in which 30% of the American population lives in New York.

[Kamala-Tim 2024]

The estimation for LGBTQ people is probably closer to reality nowadays than the “true proportion”, tbf.

[hurricanehink]

The estimation for LGBTQ people is probably closer to reality nowadays than the “true proportion”, tbf.

Came here to post this. I imagine you’ve seen the polls saying 20something percent of Gen Z is LGBTQIA+2. I’m all about that. And lolwut to the people believing 30% of Americans live in NYC. That can’t be right.

[Drop Billionaires, Not Bombs]

Love the world in which 30% of the American population lives in New York.

Or a combined >90% of the American population living in either New York, California, or Texas. No wonder this country is behind in math.

[President Johnson]

Wow, interesting Americans understimate the size of minorities. In Germany it's the opposite. I saw a poll here that people on average thought 21% of the population is Muslim, when it's indeed just about 5%.

[Zohranism is OUR future]

People in general tend to overestimate the small and underestimate the big. This applies to all number by the way. This is not shocking.

[kwabbit]

I’m pretty sure this study was ruined by the fact that the slider was hard to move so people gave up and left it near 50. That’s the explanation I’ve read. There’s no way most of these are representative.

[Vosem]

Wow, interesting Americans understimate the size of minorities. In Germany it's the opposite. I saw a poll here that people on average thought 21% of the population is Muslim, when it's indeed just about 5%.

Americans overestimate the proportion of the population which belongs to minority groups, as I think literally every country in the world does. Haven't really looked into this deeply so I would be fascinated if somebody has a counterexample, though.

[2,284,967]

Well, it's all the stuff the media talks about. Probably has an amplifying effect.

[Kabam]

Well, it's all the stuff the media talks about. Probably has an amplifying effect.

The media needs to talk less about left handed people.

Something with this poll must have gone wrong. It is probably true that the size of minorities gets overest, but some of this results are just absurd.

[2,284,967]

Well, it's all the stuff the media talks about. Probably has an amplifying effect.

The media needs to talk less about left handed people.

Something with this poll must have gone wrong. It is probably true that the size of minorities gets overest, but some of this results are just absurd.

It's absurd, yes, but it works with the idea that these are things the US media talks about 24/7  and some of these things are key issues in current day politics. I'd say that about 75% of the overestimation is explained by that and the remaining quarter is simply just people guessing.

[Darthpi - Crush the Oligarchy]

I think my favorite thing here is Muslim, Jewish, Atheist, and Catholic combining for 131%.

[Fmr. Gov. NickG]

This has been true for a long time across many countries about many minority groups.

Whenever I taught Intro to US Government (a class with ~300 students), I would survey them about the size of some of these groups.  They consistently averaged saying 30% of the US population was Black and 10% was Muslim.

I also compared this to a compilation of surveys in several European countries from about 10 years ago. Every country dramatically overestimated the Muslim population except one. The exception of course was Turkey, where 97% of the population is Muslim but its citizens significantly underestimated it.

[Nyvin]

Thinking 30% of the country lives in NYC has to be the dumbest.

[Trump v. Wong Kim Ark]

I think my favorite thing here is Muslim, Jewish, Atheist, and Catholic combining for 131%.

Yeah, that almost seems like reason enough to throw it out due to methodological issues.

[The Undefeatable Debbie Stabenow]

I think my favorite thing here is Muslim, Jewish, Atheist, and Catholic combining for 131%.

Yeah, that almost seems like reason enough to throw it out due to methodological issues.

That's not a methodoogical issue. The entire point is that people think that these minority groups are impossibly large.

[Secretary of State Liberal Hack]

I remember reading a post that stated the methodology of the study skewed the results in favour of everything being around 50%. Those surveyed were paid for their time, and the percentages were input via a slider that defaulted at 50%.

So a large proportion of those who took the survey zoomed through the sliders, moving them a little from 50% in order to finish the survey quickly instead of giving the questions serious thought.

[The Undefeatable Debbie Stabenow]

I remember reading a post that stated the methodology of the study skewed the results in favour of everything being around 50%. Those surveyed were paid for their time, and the percentages were input via a slider that defaulted at 50%.

So a large proportion of those who took the survey zoomed through the sliders, moving them a little from 50% in order to finish the survey quickly instead of giving the questions serious thought.

Now THIS means there was a methodological issue.

[インターネット掲示板ユーザー Punxsutawney Phil]

I remember reading a post that stated the methodology of the study skewed the results in favour of everything being around 50%. Those surveyed were paid for their time, and the percentages were input via a slider that defaulted at 50%.

So a large proportion of those who took the survey zoomed through the sliders, moving them a little from 50% in order to finish the survey quickly instead of giving the questions serious thought.

Now THIS means there was a methodological issue.

Sad that a potentially useful poll could be ruined by a methodological issue like this. They should have set the bars at 0% instead of 50% as a default.

[Mercenary]

True #s make sense. The estimated #s show that people are morons.

[LostFellow]

Over $100000 at 38% while over $1 million at 20% is certainly interesting. Somehow the average American is numerically illiterate enough to think that there are more households over $1 million income than between $100000 and $1 million. Either that, or there is the hidden bias of people just leaving the slider at 50%, as kwabbit and others mention, which explains the skew.

[DrScholl]

All surveys have errors, but I find it hard to believe that the results as a whole are skewed because of people not understanding how to use the slider. For all the complaints you hear about cities having too much control and rural areas having none, these numbers do support why some have those beliefs. New York City is large, but only a slim percentage of the overall population. The average person probably isn't going to properly calculate that out of a total population of 330 million percentage wise. Large whole numbers are going to equal a big percentage for some people.