科學(xué)一刻 確認(rèn)偏誤
Here’s a simple psychology experiment: Imagine four cards on a table. These cards all have letters on one side and numbers on the other. The sides you can see read A, D, 4 and 7.
That’s all you have to picture: A, D, 4, 7. Now, someone tells you that if a card has a vowel on one side, then it must have an even number on the other side. Which cards would you turn over to decide whether that is true?
Most folks, when presented with this question, choose to flip the A and the 4. After all, A is a vowel, so you want to know whether it has an even number on its back. Four is an even number, so you want to know whether it has a vowel on its back. That makes sense. But you had four choices: A, D, 4 and 7. Do either of the other two choices matter?
Try Again
Sure, the 7 does. Seven is an odd number, so if it has a vowel on its back, you can stop right there—the claim is wrong. But nobody turns over the 7. Why? Psychologists call this “confirmation bias.”
Our brains seem to be wired to look for confirming instances—that is, we get an idea of how the world is, and then try to find examples that confirm our idea. We don’t look for examples that disconfirm our idea, even though that’s just as useful, sometimes even more so.
