5 ms·
The question they asked is difficult to interpret, i.e. hard to translate between specific answers "I think the top quintile possesses N% of wealth" and an intu
by waqf 16y ago
The question they asked is difficult to interpret, i.e. hard to translate between specific answers "I think the top quintile possesses N% of wealth" and an intuitive understanding of how skewed (or not) a distribution we're talking about.
I have a strong quantitative background, and I have to do some calculations to interpret these numbers. And these guys are interviewing average Americans and inviting them to pull numbers out of their ass. Cognitive biases are going to skew these numbers horribly: people will sooner say 80% because it "sounds" like a much more "reasonable" answer than 99.8%, even if a little math would show that it wasn't.
For example, I'd like to see the study done again with people asked the mathematically almost-equivalent question: what do you think is the average wealth of the top (resp. second, etc.) quintile. (Almost-equivalent, because you have to tell the respondents the total amount of U.S. wealth for them to be able to convert one answer to the other.) Without that control I think the results are worthless.
- jfager 16y agoCognitive biases are going to skew these numbers horribly I guess I thought the point of the article was to demonstrate exactly this skew. Where does a person's emotional reaction to tax cuts or welfare come from? If people casually think that poor people aren't as poor as they really are and that wealthy people aren't as wealthy as they really are, they're going to act accordingly when it comes up in their daily life. How they actually think when forced to confront real numbers seems less relevant in that case - how many people actually sit down and do the math before deciding who to vote for?
- waqf 16y agoYou missed my point, or at least part of my point. The point is that the numbers people write down in this survey will not accurately represent their own impression of how poor poor people are.
- jfager 16y agoI don't think I missed your point - you're saying that people confronted with a question they aren't qualified or interested in enough to answer thoughtfully will be biased towards responses that "sound good". I have no disagreement there, I'm just saying that it's still interesting to know what that incorrect "sounds good" answer is in this case, because it reflects how people casually regard an issue. My unsubstantiated hypothesis is that this casual regard may actually be more important anyways, because many people aren't thoughtful in their regular daily lives, either. In other words: if a cognitive bias can affect how someone responds to a survey, why can't it affect how that same someone acts in their daily lives?
- tibbon 16y agoHowever, I think that anyone who understands general statistics and numbers as they apply to social groups would guess at this trend almost immediately as things like this almost always fit a Lorenz/Gini curve, although I guess the harshness of the reality of it really is shown here. I remember when studying some social network stuff the other year I found that pretty much everything in every social network I looked at followed this type of trend regardless of what the 'wealth' was (twitter followers, youtube subscribers, flickr views, etc)
- moultano 16y agoYou're entirely right. I dont think anyone but a demographer has an intuition for stats like this. I'd bet that a ton of people answered 20% across the board.