half_full smallA recent Time Magazine article talks about the Optimism Bias -- the phenomenon of people being eternally optimistic about the future. This can be shown in various ways including how people wildly overestimate things like personal life expectancy, marriage solvency, career prospects, etc. In this context "bias" doesn't mean something inherently bad. There is good reason to have an Optimism Bias as it leads to concrete positive outcomes like better health and longer life, not to mention inspiring us and making life generally better.

The Time Magazine article (actually an excerpt from the book The Optimism Bias by Tali Sharot) looks at this from a neuroscience perspective. When people are asked to imagine specific events in the future the general tendency seems to be to look at things through rose colored glasses. Corresponding parts of the brain light up in an fMRI machine indicating that people are indeed experiencing pleasant feelings even when the things they have been asked to imagine are quite mundane.

What does it mean for researchers? In particular what happens when we ask people about future behavior? Say, their likelihood to buy a product in the next one year.  Wouldn't the Optimism Bias mean they are likely to overstate their likelihood to buy something? This may vary depending on the kind of product (and perhaps even person), but shouldn't we expect to see it at the overall level of the sample? If so, shouldn't we apply an Optimism Discount to their purchase intentions score?

Okay, but how do we measure this Optimism Discount? Here's a thought. First ask people how likely they are to buy something in the next one year. Then one year later (when they have forgotten all about this survey) ask them how likely they would have been to buy the same thing if they had been asked a year ago. If their scores are lower, then the difference is your Optimism Discount!

Alternatively, ask one group of people about the purchase in the next one year. For another group ask them to travel back in time one year and predict their prediction. So let's say you are now in June 2011. Imagine you were in June 2010. What would be your prediction of your purchase intention of product X in June 2011? Given that they are actually living in June 2011, their prediction should be considerably less rosy. The difference between the two groups would be the Optimism Discount. Of course, it is likely that there are other explanations for the difference in scores and will have to be ruled out by running proper experiments. We'll leave that to our academic brethren for now!