The numerical rule and the confirmation bias

There appears to be real resistance to test alternatives to an initial hypothesis. That is, instead of seeking evidence that disconfirms a hypothesis for which there is some support, people tend to seek evidence that confirms this support – that is they exhibit something known as a confirmation bias.

What the 246 puzzle shows is that confirmation bias permeates our approach to trying to make sense of the world. Some have even gone so far as to argue that the confirmation bias is so powerful that it even crops up in scientific thinking. That is, once a theorist has evidence in favour of their own pet hypothesis, the inclination is to garner further support for this rather than engage in thinking about how such an hypothesis might be disproved. Try considering alternative hypotheses and testing these!

The data from the original study showed that out of 29 students tested – 22 announced one incorrect rule, 9 then went onto to announce a second incorrect rule and, of these, a further two announced a third incorrect rule. Only 6 of the original 29 announced the correct rule without first setting out an incorrect rule (and by the way the rule is any three numbers in ascending order – and please let’s try and keep this amongst ourselves so that others can engage with the task without knowing the rule!).

Additional Video Link

Here we provide an example of someone engaging with the 246 puzzle.