Thursday, January 29, 2009

Choice Blindness – What are we forgetting?

Photographer: Mark Hanlon

MIT is currently in its Independent Activities Period (IAP). Each January students take a break from regular class work and experience a variety of mini courses ranging from weighty subjects like “Energy Storage Solutions” to just for fun topics like “Build Your Own Electric Guitar.” IAP is a great way to brush up on skills or try something completely new. Fortunately for me the non-credit courses are open to alumni so I’ve spent my week immersed in “Statistics and Visualization for Data Analysis and Inference” (very useful but not much fun) and “Philosophy of Cognitive Science – Choose your own adventure!” (esoteric but mind-bendingly interesting).

Yesterday in the cognitive philosophy course we covered research by Petter Johansson and his colleagues from Lund University. In a 2005 Science paper entitled “Failure to Detect Mismatches between Intention and Outcome in a Simple Decision Task,” the authors present a series of experiments which lead to a construct Johansson calls Choice Blindness. As stated in the paper’s introduction:

“A fundamental assumption of theories of decision-making is that we detect mismatches between intention and outcome, adjust our behavior in the face of error, and adapt to changing circumstances. Is this always the case? We investigated the relation between intention, choice, and introspection. Participants made choices between presented face pairs on the basis of attractiveness, while we covertly manipulated the relationship between choice and outcome that they experienced. Participants failed to notice conspicuous mismatches between their intended choice and the outcome they were presented with, while nevertheless offering introspectively derived reasons for why they chose the way they did. We call this effect choice blindness.”

It is worth taking a more detailed look at the experiment to grasp the full magnitude of this surprising result. Subjects (male and female) were presented with pictures of two different female faces at a time on two separate cards. They were then were asked to point to the picture they found the most attractive. In the manipulation condition, the experimenter performed a slight of hand card trick and presented back to the subject the card they did not pick and asked them to justify their choice. In most cases these subjects proceeded to justify why they choose this other woman as if it was the card they had intended to choose all along, completely unaware of the fact that they DID NOT choose this picture. Amazing!

Why does this happen? The authors suggest that subjects fail in introspection. Johansson believes that, at the time of outcome, subjects no longer have access to their original intentions. A “fundamental assumption of theories of decision making [that] intentions and outcomes form a tight loop” is somehow broken.

There may be another, somewhat complementary, explanation in “motivated forgetting.” Motivated forgetting suggests that given a strong motive and suitable vehicle for belief, people are capable of forgetting information that does not suit their motive and falsely remembering alternative “facts” that do. Note that motivated forgetting involves the subject authentically forgetting the true facts and not just conveniently pretending to forget them for a social/external benefit. In the case of the card experiment, subject desire for consistency may be a strong enough motive to cause subjects to forget their original intent and remember it as it’s opposite.If motivated forgetting is playing a role one might predict that subjects faced with a different motive would not exhibit choice blindness. The following experimental idea could use some work but what if subjects were evaluating and choosing between two race horses before a race at the betting window. The experimenter would then place a bet on the behalf of a subject based on the subject’s choice. In the manipulation case experimenters would ask for a bet, making sure the subject overheard them, to be placed the opposite horse to the one the subject chose. If motivated forgetting is playing a role, I would predict that when the false choice horse won subjects would not remember that this horse was not their original choice and be able to explain the reasons they “chose” this false horse much like in the Johansson experiment. An important difference would arise when the false choice horse lost and the true choice horse won. In this case subjects may remember having actually chosen the winning horse as their motive is now different. This second result would not be predicted with choice blindness.

Wednesday, January 7, 2009

Scapegoat the Visual

I have a meeting with Harvard Professor Max Bazerman today. As most readers will know, Professor Bazerman is world renowned for his research on negotiations and judgment in managerial decision making. He also studies bounded awareness and ethicality, societal decision making, and want/should conflicts. Along with collaborators Don Moore, Francesca Gino, and Lisa Shu, Professor Bazerman’s recent research dives into several aspects of “Moral Luck” discussed previously. In preparation for the meeting I created the following visual to better explain some of my thinking on the topic. I am happy to share it here on

Saturday, January 3, 2009

BITH: Buying Behavior Part Two

This year’s Society for Judgment and Decision Making conference featured a symposium organized by CMU’s Leslie John and Jessica Wisdom on “Behavioral Economics and Health.” One of the papers presented in this symposium (by Ms. John) just became Behavior in the Headlines (BITH). The study, “Financial Incentive–Based Approaches for Weight Loss,” was recently published in JAMA and it is starting to be picked up by the popular press including the Pittsburgh Post-Gazette. The CMU website sums up the results quite nicely:

“[The study] …placed adult dieters into three groups. One group entered a daily lottery and received winnings only if they reached their targeted weight levels. A second group invested their own money, but lost it if they didn't meet their goals. The third group was given no monetary incentive at all.

The goal: lose a pound a week over 16 weeks.

The results were striking. The mean weight loss for both incentive groups was more than 13 pounds — with about half the participants reaching the 16-pound goal. But the mean weight loss for the control group was only 4 pounds.”

Followers of this blog will note that this is not the first time that schemes using financial rewards to elicit good behavior and their potential pitfalls have been discussed. I am curious as to how these results square with Uri Gneezy and Aldo Rustichini’s study mentioned in previous posts on an undesirable practice of day care center parents. In their study, when financial incentives were used to incent good behavior it backfired in an irreversible way. Dan Ariely’s explanation claims that when incentives are taken from the domain of social exchange to the domain of economic exchange the process cannot easily be reversed. This would suggest that when the financial incentive is removed in the weight loss study, subjects should whiplash back to bad behavior, as they have no further economic motivation to continue. What happened?

During the next seven months the subjects in all groups did gain weight back but apparently not as much weight as they had lost. To me this result is inconclusive. Logically, if a lot of weight was lost it will require that a lot of weight to be added back, which simply takes time. Perhaps seven months is not long enough. How rapid was the rate of weight regain by subject group? Are the high weight loss subjects gaining back at a faster rate? Will subjects eventually overshoot their old weight and end up heavier and worse off than they were before? The next layer of questions involves whether or not it was the financial incentive itself or the focusing/scoring/gaming process (beyond mere weigh-ins) that caused subjects to lose weight. In summary this is a great study attacking an important problem while opening up a number of good new research questions. I’m "hungry" for more papers from this world class group of collaborators.

Co-Authors: Kevin G. Volpp, MD, PhD; Leslie K. John, MS; Andrea B. Troxel, ScD; Laurie Norton, MA; Jennifer Fassbender, MS; George Loewenstein, PhD