In his book Misbehaving: The Making of a Behavioral Economist, Richard Thaler talks about his earliest days of collaboration with Daniel Kahneman and Amos Tversky, who were already giants in the field of cognitive psychology. In their 1974 paper “Judgment Under Uncertainty,” Kahneman and Tversky took the academic world by storm, having proposed three straightforward cognitive biases of human decision-making. Over the next few years, their further work on choice, risk, and uncertainty revealed even more anomalies in decision-making, which had such a powerful effect on other academic disciplines that in 2002 Kahneman won the Nobel Prize In Economics.
All of a sudden, people were paying attention to cognitive bias like never before. Part of this involved rediscovery and renewed attention to some of the facts about human psychology that were so old no one could really be sure who had first discovered them. But along with these came fresh work that revealed a number of other built-in cognitive biases. And two of the most important for our purposes build on Peter Wason’s earlier discovery of confirmation bias. These are the “backfire effect” and the “Dunning-Kruger effect,” both of which are rooted in the concept of motivated reasoning.
Motivated reasoning is the idea that what we hope to be true may color our perception of what actually is true. We often reason, that is, within an emotional context. This is arguably the mechanism behind the ideas of dissonance reduction and confirmation bias, and it is easy to see why. When we feel psychic discomfort we are motivated to find a non-ego-threatening way to reduce it, which can lead to the irrational tendency to accommodate our beliefs to our feelings, rather than the other way around. Upton Sinclair perhaps said it best when he observed that “it is difficult to get a man to believe something when his salary depends upon him not believing it.”
The idea of confirmation bias seems straightforwardly related to motivated reasoning in that it is customarily when we are motivated to defend the idea that one of our beliefs is right that we look for evidence to confirm it. We commonly see this mechanism at work in police detectives, who identify a suspect and then try to build a case around him, rather than search for reasons to rule him out. It is important here, however, to distinguish between motivated reasoning and confirmation bias, for they are not precisely the same thing. Motivated reasoning is a state of mind in which we find ourselves willing – perhaps at an unconscious level – to shade our beliefs in light our opinions; confirmation bias is the mechanism by which we may try to accomplish this, by interpreting information so that it confirms our preexisting beliefs.
In his work on the psychology of emotion and moral judgment, David DeSteno, a psychologist at Northeastern University, has studied the effect of “team affiliation” on moral reasoning. In one experiment, subjects who had just met were randomly divided into teams by giving them colored wristbands. Then they were separated. The first group was told that they would be given the option of performing either a fun ten-minute task or a difficult forty-five-minute one. Each subject was then placed alone in a room and told that he or she should choose which task to do or, to be unbiased, decide by a coin flip, but in either case the person who entered the room afterward would be left with the remaining task. What subjects didn’t know is that they were being videotaped. Upon exiting the room 90 percent said that they had been fair, even though most had chosen the easier task for themselves and never bothered to flip the coin. But what is absolutely fascinating is what happened next. When the other half of the subjects were asked to watch a videotape of the liars and cheaters, they condemned them – unless they were wearing the same color wristband. If we are willing to excuse immoral behavior based on something as trivial as a wristband, imagine how our reasoning might be affected if we were really emotionally committed.
コメントを残す