"We are all average"

One of the most ubiquitous cognitive biases is overconfidence. As Daniel Kahneman writes:

The observation that “90% of drivers believe they are better than average“ is a well-established psychological finding that has become part of the culture and it often comes up as a prime example of a more general above-average effect. (Thinking, Fast and Slow, p259-60)

However, this is not the only cognitive bias that arises when people are asked to estimate their position on a statistical distribution. A new paper by Daniel Sgroi and Eugenio Proto describes another one - that “individuals tend to see themselves as more “average" than is the case.":

Those at the extremes tend to perceive themselves as closer to the middle of the distribution than is the case. For example those who are left-wing see the population as more left-wing and feel themselves to be more typical, and those who own a particular type of mobile phone are likely to misperceive their own brand as more popular than is the case…taller and heavier individuals think that there are more tall and heavy individuals in the population.

This is not always a self-serving bias, nor is it the same as overconfidence; tall men, for example, would probably feel better about themselves if they had a more accurate image of their superior height.

The explanation for this might be straightforward. Likes tend to attract likes - lefties associate with lefties and lardies with lardies - and this, combined with the availability heuristic, leads people to over-estimate the extent to which others are like them.

I suspect this bias might have significant social significance. I can think of two ways, and there are probably more:

1. It leads to the “middle England error” - the tendency of rich people to assume that their incomes are more typical than they really are. This can in turn cause the rich and influential to under-estimate the depth and breath of poverty.

2. Criminals often claim that “everybody’s doing it”, perhaps because many of their friends are.  In believing that behaviour is normal, though, they are more likely to do it. And so crime is higher than it otherwise would be.

Comments

[Luis Enrique]

this reminds me of that:
http://flipchartfairytales.wordpress.com/2012/03/01/how-normal-is-your-social-circle/

[Rick]

Thanks Luis, you beat me to it!

[Niklas Smith]

So true! This post reminds me of the last Sunday Times before the 2010 general election. The main leader and a double-page spread were devoted to trashing the Lib Dems' economic and particularly tax proposals, from which one thing jumped out at me: the article said that abolishing higher rate tax relief on pensions would hurt "the middle class". Given that only 3.8m out of 31.7m income taxpayers (12%) paid higher rate tax in 2010 that is a very odd definition of the middle class!

Source for that statistic is this must-read Economist article about how the middle class is misinterpreted in British politics, and how the real middle class have modest incomes and certainly don't send their children to private schools: http://www.economist.com/node/15777629

[james higham]

The thing is - some people just are better, mediocritizing "studies" notwithstanding. It makes the world go round.

[Simon Jester]

The thing is, if 90% of drivers believe they are above average, then a majority of them probably are. (Assuming that the other 10% are correct in believing themselves below average, and that the average is a threshold, not a band.)

[Left Outside]

"You're all individuals"

"I'm not"

[Francesca]

I don't think this counts as a social effect, but one of the consequences of this that I notice is that political campaigning is made more ineffective. I'm thinking in particular of the AV campaign, which was slated by someone as being "by Guardian readers for Guardian readers", most of whom were always going to vote for AV anyway. The campaign was very unsuccessful because its (highly educated, very liberal) designers assumed that the general public was just like them, and would be influenced by the arguments that drew them to the issue. Of course, this turned out not to be true.

It's why I am very suspicious of Twitter. I think it's an echo chamber, because people follow and are followed by like minds, but it gives us a fallacious impression that this is what the world is like.

Also, I wonder whether it increases suspicion of difference - that we are more inclined to ascribe hostile or immoral motives to people who are different from us, because they seem to us to be deviating from a norm that we made up? I can't adduce any evidence for that, though.

I remember the first Conservative I met when I was in college. It took me several days to believe that he really was both a Conservative and a student, and he wasn't just having me on.