Consider these three scenarios:
- An investor decides to try a brand-new strategy trading contracts in the futures market. After five straight winning trades, he declares he’s found the key to success.
- A schoolteacher is convinced that a particular student is going to struggle with a new unit. The next day, she asks the student two especially difficult questions; when the student answers them incorrectly, she decides that he really is struggling and makes a note to give him extra help.
- A scientific researcher forms a hypothesis and designs an experiment to prove it. After a few trials that are consistent with her prediction, she concludes that her theory was indeed correct and tries to publish her findings.
In each case, someone fell prey to confirmation bias.
Confirmation bias is simply the human tendency to focus on evidence that corroborates, or confirms, a prior assumption or belief. It’s an understandable mistake. It’s also potentially destructive.
Mental Explanations for Confirmation Bias
Psychologists are confident that confirmation bias exists – after all, it’s been an observed phenomenon for thousands of years – but they’re still divided on its cause. Some believe it’s just wishful thinking, a natural tendency to look on the bright side and favor pleasant rather than unpleasant thoughts. We subconsciously discard information that is somehow unfavorable and focus on the things we’d prefer to be true.
A related explanation focuses on the availability heuristic, a mental shortcut that causes us to only use the information that’s immediately at hand to judge a situation. Since facts that meet our original expectations are easier to recall, we give them much more weight in our thinking.
One more economic explanation focuses on cost-benefit analysis. Some psychologists theorize that instead of seeking truth, the brain is naturally wired to avoid costly errors. Since accepting a false hypothesis is often less damaging than rejecting a true hypothesis, the theory goes, we seek out information that confirms rather than challenges our hypotheses.
Regardless of the cause, confirmation bias is insidious. It affects everyone’s thinking to some extent, and even when we’re aware it exists, it can be tough to shake.
The Dangers of Confirmation Bias
Let’s return to our first example. Under normal circumstances, a rational investor would realize that it’s unwise to make a trading decision based on just five trades – that could just be a lucky streak. Because he was expecting his strategy to work, however, this investor ends up giving undue weight to a small sample size. He might decide to implement his new strategy too aggressively and end up losing money.
Our second example, the schoolteacher, is also guilty of using a small sample size. In this case, however, the teacher is doing something particularly problematic: specifically looking for information that confirms her bias. By asking difficult questions, she consciously or unconsciously sets up the student to fail.
In the third example, our scientist has committed a grave error by failing to seek out falsifying as well as confirming evidence. In addition to trying to show her hypothesis to be correct, she ought to have designed an experiment that would challenge her initial assumption.
Confirmation bias can be even more subtle. Often, people who do seek out alternative explanations or opinions go in with the mindset that they’ll just prove the other side wrong. That leads to selective listening: focusing on the flaws in the opposing argument without considering the strengths.
Dealing with Confirmation Bias
It’s practically impossible to shake off confirmation bias entirely, but there are ways to cope. One method is to seek out objective measurements. Data that can be interpreted subjectively is naturally prone to bias. Numbers are not.
Second, use a larger sample size whenever possible. The less information you have, the more likely you are to rely on your assumptions and biases to fill in the gaps. Bringing more information to light, especially from multiple sources, limits the power of confirmation bias.
Third, explicitly seek out dissenting opinions and approach with a fair mind. Again, there’s no way to keep confirmation bias from creeping in at all, but trying to consider the other side’s views as calmly and rationally as possible will help.
Fourth, avoid making decisions when at emotional extremes. This is a good mantra in general, but it’s relevant particularly to confirmation bias. Assumptions are usually rooted in emotion, not fact, and they have more power when we’re especially happy or especially unhappy.
Finally, and perhaps most importantly, don’t make decisions alone. Bringing in multiple people with multiple assumptions tends to negate the power of each individual assumption. As long as every member of the group doesn’t have the same expectation, the members’ respective biases will effectively cancel each other out.
As great as it would be to go into situations perfectly objectively, with no preconceived notions, the human mind just doesn’t work that way. Confirmation bias is always a concern, but by proactively recognizing its influence, it’s possible to limit its impact.