In that vein, there was an intruiging article in the Economist this week that reported on a conjoint study that showed just how stark the difference can be (see here). In summary though a test was done to assess students team mate choice for a quiz in which they were told they would win money. They had information on IQ, experience in quizzes and education level. They also had a photo that happened to show how fat or thin the person was. Afterwards they were asked what attributes they used to choose, and they all reported that weight was the least important attribute … but the data in fact showed that it was the most important. In fact, they would give up half (11) of the range of IQ scores to get a thin person.
I am trying to work out what heuristic might make this feel right, and can only come up with one that says thinner people are likely to have a faster metabolic rate than fatter people, and this just might make a difference to their reaction time if not mental agility. So, all else being equal it might just be valid to choose a thinner person (if not very socially inclusive). However, it would seem a poor model for choice when you have much better data at hand.
What is really stark to me is not that people are making a choice based on apparently dodgy data, but that they are completely unaware that they are doing it … indeed they believe that they are doing the opposite. And, worth recalling that these are not educational drop-outs, they were students. This always seems the sad bit to me – not that we have heuristics for making decisions that can fail us from time to time (usually called cognitive biases), but that we are often aware of the types of failure, and so make them time after time in blissful ignorance.