The Better-than-Average Effect

A 1981 study by psychologist Ola Svenson (PDF link) asked two groups of participants, Swedish and American, to compare their driving skills to those of their peers. They found that 93% of the US drivers and 69% of the Swedish drivers believed themselves to be more skillful than the median driver in their group.

Now it's theoretically possible for more than 50% of a group to be better than average at something, if average is taken to be the arithmetic mean. But the Svenson study was about the median, so the participants' beliefs can't possibly be true. This phenomenon, known as the better-than-average effect (BATE), has been replicated in multiple studies, and in many areas of life.

But I think this particular form of irrationality isn't limited to believing one is better than average. In my (completely anecdotal) experience, it also extends to some value-neutral domains (where there is no obvious better or worse), and at least one domain where I suspect most people believe they are worse than average.

Procrastination, I think, is an example of the latter. My guess is that most people believe they are worse than the median (i.e. that they procrastinate more than average).

I suppose you could turn it around and argue that this is just another manifestation of BATE (people think they are better than average at procrastinating). But that's just a matter of framing, and if it comes to that, most questions could be similarly flipped. For example, instead of asking drivers if they are more skillful than the median, you could ask if they are more likely than the median to cause an accident. In the second case, my guess is that most drivers would say they are less likely than the median.

BATE typically skews towards positive self-evaluations – in fact the article I linked to earlier defines it as "the tendency for people to perceive their abilities, attributes, and personality traits as superior compared with their average peer" – and procrastination is generally considered a negative trait. So if my hunch is right, procrastination is an exception: an example of a worse-than-average effect.

Now for the two value-neutral examples.

Consider the question of how strongly you feel versus how much you show. We might call this trait emotional demonstrativeness. For example, Chris Evert, in the passage I quoted in this post, was suggesting she was less emotionally demonstrative than Goolagong. If you did a survey, I reckon you'd find that most people think they are less demonstrative than the median.

Of course, how strongly you feel is a subjective state, so it's impossible to empirically compare emotional demonstrativeness. As I wrote in that post, "Presumably Goolagong reacted more vehemently than Evert did when she missed a volley. But perhaps Goolagong really did feel the disappointment more keenly – who can say?"

My other value-neutral example, however, is empirically testable. Variability in human attractiveness to mosquitoes can and has been studied and compared. Nevertheless, if you did a survey, I think most people would say they are more attractive to mosquitoes than the median.

Kolkata, where I grew up, and Singapore, where I live now, both have lots of mosquitoes, and over the years, I've heard many people say they are unusually attractive to them. I recently went hiking with a friend, and she said (unprompted) that she is less attractive to mosquitoes than average. As far as I can remember, it's the first time in my life that I've heard anyone say that.