Tips on Arguing: Confirmation Bias

Inquiry:

Whether or not you’ve ever lived with someone of the opposite sex, you’re probably familiar with the following scenario:

Maria storms into the living room, where she finds Norm, her husband, watching TV.

“You forgot to put the toilet seat down,” she accuses, her eyes narrowing to narrow slits. “I almost drowned myself.”

Norm rolls his eyes in response. “C’mon,” he says, waving her away with his hand. “So I forgot this once.”

“You always forget!” Maria counters.

“I do not,” he replies. “I remembered every other time this week.”

In the above situation, both Maria and Norm suffer from confirmation bias: the tendency to interpret data based on the parts of it that fit most favorably with our beliefs. Maria sees when Norm leaves the seat up, but she can’t tell when he puts it down. Her memory is filled with her husband’s mistakes. By the same token, Norm recalls all of the times that he thought to put the toilet seat down. But he can’t possibly remember the times when he forgot to put the seat down; if he had, he wouldn’t have left it up. Norm’s memory is filled with examples of his own successes.

According to the Skeptic’s Dictionary, a book and online reference by writer and retired professor of philosophy Robert T. Carroll, Ph.D., confirmation bias is an effect of human psychology. The mind has an easier time building upon preconceived notions than tearing them down. “The tendency to give more attention and weight to the positive and the confirmatory has been shown to influence memory,” he says. “When digging into our memories for data relevant to a position, we are more likely to recall data that confirms the position.”

Although putting the toilet seat down might not be a huge deal, the design and interpretation of scientific experiments is utterly dependent on evading the threat of confirmation bias to the greatest extent possible. One way that researchers handle the problem is by introducing randomized samples into their tests. This keeps them from unwittingly selecting a group of people with a shared characteristic. Another method is by double-blinding a trial; neither the subjects nor the researchers know whether a person belongs to the control group or the experimental group.

Even after an experiment is finished, groups of outside scientists will peer-review the report of what happened. This process helps to ensure that most biases get caught. Not even science, however, can completely escape confirmation bias. That’s why no single study is ever absolute – it must be repeated over and over by others before it can be accepted.Although no one can ever completely eliminate bias, you can mitigate it simply by considering the reasons a person might have for maintaining the opposite viewpoint. It requires more work then you’d think, but the results are worthwhile: your own arguments become more robust, and your mind becomes more open.