Elliot comments on The Halo Effect - Less Wrong

23 Post author: Eliezer_Yudkowsky 30 November 2007 12:58AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (49)

Sort By: Old

You are viewing a single comment's thread.

Comment author: Elliot 30 November 2007 03:05:44AM 4 points [-]

For example, a study of the 1974 Canadian federal elections found that attractive candidates received more than two and a half times as many votes as unattractive candidates (Efran & Patterson, 1976).

This does not mean what it sounds like it means. Well, it *could*, but it doesn't have to. Specifically, this result is consistent with the voters' claims that they don't vote for candidates because of physical attractiveness.

This is a case of "correlation does not imply causation". Just because good looks were correlated with votes doesn't mean they caused votes. There could be another effect causing both.

Such effects are easy to imagine. For example, perhaps people with good looks receive more encouragement in school and from their parents, and thus turn out smarter. Then they could have received all those votes because they were genuinely better candidates. This particular possibility may have been looked for and ruled out, but there are infinitely many others.

The important thing is that you can't find the truth purely by finding correlations. What you need are explanations. Specifically, there needs to be a detailed explanation of how being more attractive *causes* favoritism (and also of what causes people to be blind to their own favoritism). And when we have that explanation, then we can compare it to rival theories that explain the observed data, including the correlation, in other ways.