Confirmation Bias: Seeing What We Want to See
We give more weight to evidence that confirms what we already believe and discount evidence that contradicts it. This is the most pervasive cognitive bias in everyday reasoning.
Confirmation bias is the tendency to give more weight to evidence that confirms our existing beliefs and to discount, ignore, or rationalize evidence that contradicts them. The bias is pervasive, it is automatic, and it operates below the level of conscious awareness most of the time.
The classic study, by Peter Wason in the 1960s, asked subjects to test a rule. The rule was simple: the experimenter would offer a sequence of three numbers, and the subject had to figure out the underlying pattern. The actual rule was “any three numbers in increasing order.” Subjects could test their hypotheses by proposing new sequences and being told whether they fit the rule.
Most subjects guessed something more specific — “numbers separated by 2,” for instance — and then proposed sequences that fit their guess. The sequences did fit the rule, which subjects took as confirmation. Few subjects tried to disconfirm their hypothesis by proposing sequences that violated it. The bias toward confirmation was robust across subjects, education levels, and intelligence.
How it shows up
Confirmation bias shows up in everyday reasoning in several ways. We pay more attention to news stories that confirm our political views. We remember anecdotes that support our positions and forget the ones that don’t. We give friends a charitable interpretation and skeptics a harsh one.
In each case, the bias is not that we’re lying or being deliberately unfair. It’s that the cognitive machinery that processes evidence is asymmetric. Confirming evidence is processed easily, accepted, and integrated. Disconfirming evidence is processed slowly, scrutinized harshly, and often rejected.
Why it persists
Several reasons. First, it’s emotionally easier. Updating beliefs in light of contradictory evidence is psychologically costly. Confirming beliefs is comfortable. Second, it’s socially advantageous in many contexts. Being consistent — sticking to your views — signals reliability. Updating publicly can be read as weakness, even when it’s exactly the right move.
Third, confirmation bias often produces correct results. If your existing beliefs are good ones, evidence that confirms them is more likely to be correct than evidence that contradicts them. The bias has a kind of working logic in environments where you start with mostly-correct beliefs.
The trouble is that we apply the same processing to environments where our starting beliefs are wrong. The bias is indifferent to whether the starting belief was correct. It just defends whichever belief is there.
How to mitigate it
Standard advice: deliberately seek out evidence that would contradict your current view. Read sources you disagree with. Imagine the strongest argument an opponent would make. Ask yourself, “what would change my mind?”
The last question is the most useful. If you can’t name evidence that would change your mind, you are probably not holding the belief for evidence-based reasons. You are holding it for other reasons — identity, loyalty, comfort — that are not themselves illegitimate but that are different from evidence-based belief.
The honest answer to “what would change my mind” is often: less than I’d like to admit. Recognizing this is the first step toward being a more careful thinker.
