In a fascinating column, John Tierney writes:
The Yale experiment was a variation of the classic one that first
demonstrated cognitive dissonance, a term coined by the social
psychologist Leon Festinger. In 1956 one of his students, Jack Brehm,
carted some of his own wedding gifts into the lab (it was a low-budget
experiment) and asked people to rate the desirability of things like an
electric sandwich press, a desk lamp, a stopwatch and a transistor
radio.Then they were given a choice between two items they
considered equally attractive, and told they could take one home. (At
the end of the experiment Mr. Brehm had to confess he couldn’t really
afford to give them anything, causing one woman to break down in
tears.) After making a choice (but before having it snatched away),
they were asked to rate all the items again.Suddenly they had a
new perspective. If they had chosen the electric sandwich press over
the toaster, they raised its rating and downgraded the toaster. They
convinced themselves they had made by far the right choice.So,
apparently, did the children and capuchin monkeys studied at Yale by
Louisa C. Egan, Laurie R. Santos and Paul Bloom. The psychologists
offered the children stickers and the monkeys M&M’s.Once a monkey was observed to show an equal preference for three colors
of M&M’s – say, red, blue and green – he was given a choice between
two of them. If he chose red over blue, his preference changed and he
downgraded blue. When he was subsequently given a choice between blue
and green, it was no longer an even contest – he was now much more
likely to reject the blue.
I would distinguish between self-deception and an endowment effect. We value more what is ours, perhaps because of our biological programming — to protect our children above those of others — spills over into decisions more generally. (Or perhaps because of a precommitment strategy to limit violent plunder of our resources.) Self-deception is then layered on top, but in fact many mothers will argue that their kids are lazier or less obedient than the average. The endowment effect holds nonetheless, as those mothers care more about their kids. It is very hard to switch back babies once the hospital makes a mistake in allocation (how much time must elapse?), even if the parents know for sure they did not take home the genetically appropriate little bundle of joy.
I can see that the monkeys behave according to an endowment effect. I am less sure that the monkeys self-deceive. The key question, in my view, is whether the monkeys would throw out or downgrade information that some other bundle of food was in fact better than M&Ms.
Great post! Now relate this concept to people’s committment to strongly held political or philosophical viewpoints, and you are really getting somewhere in explaining higher primate behaviour!
I’m puzzled. What’s silly or irrational about avoiding buyer’s remorse?
And is it self-deception or a conscious decision to move on?
Once I’ve made a finely balanced decision, I consciously shut down my brain when it tries to second-guess itself. Who wants their brain spending the next 20 years obsessing over whether it should have chosen the sandwich press or the desklamp?
And if John Tierney thinks I’m silly or pathological, I think that says far more about John Tierney’s faults than mine.
Did anyone do any pre-testing or debriefing to determine if the monkeys were confused by the experiment? Actually, I’m wondering how they got the monkeys to express an equal preference for the three colors. Given my curiousity I found a link to the actual paper, although you may need some access to the journal (my university apparently has access):
http://www.blackwell-synergy.com/doi/abs/10.1111/j.1467-9280.2007.02012.x
The 1956 study does provide a nice example for the case against deception in the lab. Number 1, you’ve reduced someone to tears which is a big no-no nowadays. Number 2, if this person came back to the lab for a different experiment after experiencing this “traumatic event” of being told she could have one of the items and then being told she couldn’t, how could we ever expect this person to truly reveal preferences, beliefs, whatever in the future?
Given that monkeys do not understand the Mars Co. commitment to making all colors taste the same, I would expect them to choose the color that most recently tasted good. Or did the experiment somehow make them choose without tasting? How do monkeys behave when you don’t give them their choice but make them choose again?
This sounds like a very difficult experiment to design.
After the monkey chose red over blue, he subsequently chose green over blue as well. What would he have chosen as between green and red? Did the original choice of red over blue lead to an upgraded opinion of red to go with the downgraded opinion of blue?
Also, why is this surprising? As others have pointed out, once you’ve evaluated a situation and made a decision, it’s perfectly sensible not to rethink the whole thing when that situation or one very similar to it comes up again.
I find the M&M experiment unconvincing. In this case, monkeys are choosing colors rather than flavors. Personally, I strictly prefer the M&Ms in red to them in any other color; and really hate them in dark brown. With a large bowl of M&Ms, I often eat the red ones first, and then the other colors except for the dark brown. When there are only those dark brown ones left, I eat them as chocolate beans.
Also, how are we supposed to receive psychological studies in which individual participants have very small compensation incentives? I saw a paper a while back by an economist at Georgia State and another from UVA (Charles Holt maybe) that used much larger payouts than Kahnmen and Tversky had used in their original studies (which, if I’m not mistaken, did not actually involve monetary payouts, but hypothetical payouts). When very large lotteries were used (in terms of money won), these economists found more evidence for VNM utility than prospect theory. That said, I’ve been skeptical of psychology experiments with weak monetary payouts ever since.
I recall the story of King Sol (or is it Solomon?). Anyway he concludeds who the “real” mother is by threatening destruction of the child. However, one of these women had possession of the child and would therefore over-value the child. The other did not have the same over-value. Is it possible that the women who had possession (and therefore over-value) would always give up that possession rather than see the child die regardless of who the real mother was?
(At the end of the experiment Mr. Brehm had to confess he couldn’t really afford to give them anything, causing one woman to break down in tears.)
Wait, were these psychiatry patients or normals?
Here’s a question for Tyler. How do the results of these two studies match with the somewhat frequent feeling at restaurants that someone else has ordered a better food choice than you when the food arrives? Is it simply a matter of having additional information that overcomes this bias? Or is there something else involved?
And is deciding that your dinner companion’s food tastes better than yours a sign that you don’t properly understand your own preferences, or is it a sign that the incomplete information in the menu did not provide the information necessary to act on your preferences?
I doubt there are many people better qualified to examine a restaurant-related question from an economics perspective than Prof. Cowen.
Relating the case of the monkeys to an economic perspective, I would have to comment that the monkeys are forced with making a choice. They face an opportunity cost when forgoing the opportunity of getting the next best thing (the blue M&M in this instance). As far as the distinguishment between self-deception and the endowment effect, I think the monkeys exhibit the endowment effect. They become reluctant to choose the color they once rejected. In essence, the color chosen does, in a way, belong to them. As with the comparison given, mothers care more about their own kids. In relation, the monkeys now value their own color of M&M’s more. Conversely, I don’t think that the monkeys show any sort of self-deception in their actions. Without the proof of the mental thinking of monkeys, though, this is hard to place as factual. In my opinion, though, when rejecting a previously rejected color of M&M the monkeys are not trying to convince themselves that they made the right choice. Most humans demonstrate a force of habit, and monkeys being our close counterpart, are more likely to do the same. Therefore, they continually reject the blue M&M.
I will raise one question, though. I am reluctant to believe that once the monkey rejected the blue M&M, and was given the option between it and another color (the green M&M), they would once again reject the blue one. If preference was given to three colors (red, blue, and green), and a choice made between the red and blue, I feel it would be necessary to give the monkeys a choice between the red and green M&M’s next. This would determine their preference for the red. Then an option between the green and blue should be given. The monkeys may simply have a preference for the red instead of a dislike for the blue. Providing this sequence to the monkeys would answer this question. It is possible that when choosing between the green and blue M&M’s, the monkeys rejected the blue simply because there was no red present to pick. The monkeys could have just as well rejected the green. I think this question would be one to experiment with before determining the endowment effect.
Comments on this entry are closed.