Confirmation bias

Confirmation bias

Most people like to think of themselves as being fairly reasonable, with their opinions and ideas the result of thought, and so open to being challenged.  Unfortunately we are all subject to “confirmation bias”, whereby we more readily look for, accept, and remember facts and opinions that confirm what we already believe.

One example of how confirmation bias leads us to look only for information that confirms our belief was described by the psychologist Peter Wason.  He showed people the sequence of numbers “2,4,6” and asked them to think what the underlying rule behind those numbers might be.  He then asked them to come up with another set of numbers to test that underlying rule.  Take a moment now to think what the rule might be, and a set of numbers to test that.

The majority of people will have come up with “3,5,7” or similar, and that also meets the underlying rule behind the “2,4,6” sequence.  However the majority of people were also unable to accept that they were wrong in thinking that the underlying rule was that the number rose by 2 each time – after all their sequence of “3,5,7” had proved it.

This is where confirmation bias can blind us to other possibilities.  The actual underlying rule is simply that the numbers increase, not that they increase by 2 each time.  A better sequence to have picked would be “3,5,8”, as this way the theory that they always rise by 2 would have been tested.  Instead we don’t just reach a conclusion but apply a test that will tend to confirm that conclusion is correct.

The way in which confirmation bias drives our willingness to accept information was shown during the 2010 general election.  Two groups of voters, one Conservative and one Labour, were shown speeches by the party leaders, given factual information about the economy and NHS, and then asked to report any changes in their opinions and give the reasons.

No member of either group fundamentally changed their beliefs, but both groups saw their members’ beliefs become stronger based on exactly the same evidence and speeches.  As you might expect their confirmation bias saw them more likely to recall and accept statements and statistics that supported their existing beliefs.  What might be more surprising was how easily members of both groups took information that clearly contradicted their assumptions and then adopted them in ways that actually confirmed them.  Essentially even presenting them with the sort evidence that they had previously described themselves as being open to considering only made their existing beliefs stronger.

How confirmation bias affects what we remember has been shown numerous times both in experiments and in real life.  Having read an even handed description of a fictional woman people were either told that she was a librarian or successful saleswoman.  Those descriptions alone were enough to influence whether people remembered information that supported her being introverted or an extrovert.  The same thing is seen in the “halo effect” in workplace appraisals, where managers’ recall of individual’s performance is compromised by their existing view of them.

There has been further research into the physiological processes that lie behind confirmation bias, with one contention being that facts congruent with our beliefs are more readily recorded.  The opposing theory, that facts that are contrary or challenging are more easily recalled has also been put forward and tested.  That neither has been proven to the satisfaction of its opponents is I suppose in its own way further demonstration of confirmation bias in action.

If you enjoyed this why not sign up for my monthly newsletter here.