Cognitive biases – distortions of reality
Cognitive biases arise from heuristics (shortcut thinking) and systematically distort the way we think and affect our decision-making. Three of the most common cognitive biases are confirmation bias, anchoring bias and cognitive dissonance. All three arise from our tendencies to focus on part of a story, especially the part that confirms our pre-existing beliefs, and to avoid holding two inconsistent beliefs at the same time. That is, we are selectively attentive, and filter out what we don’t agree with.
Confirmation bias – this is when we seek out information that tells us we were right all along! It affects the type of information we seek out, and also how we interpret neutral information that we meet along the way. If I post on Facebook or Instagram and don’t get immediate positive feedback, I feel that people dislike what I’ve posted, and are ignoring me. I think we can all identify with this. We look for evidence that supports our pre-existing beliefs about ourselves, or others, rather than useful information.
Wason (1960) demonstrated the confirmation bias by giving British university students the 3 ascending numbers 4-6-8 and asking them to guess the rule he had used to devise the series. (It was simply ‘any 3 ascending numbers’.) Students had to generate their own sets of 3 numbers, and ask if they conformed to the rule. Of course, they did, as any three numbers that were neither identical nor descending conformed, but this stopped them from isolating the rule. They were seeking evidence that confirmed what they thought, not evidence that refuted the rule, so it could be more easily identified. This is a mistake, as an attempt to refute the hypothesis would have given more clues to its true identity.
Anchoring bias – this is when we make estimates of a total or of a probability by starting from an initial idea, which ‘anchors’ us to it. An experiment described by Tversky & Kahneman (1974) illustrates this:
Two groups of high school students estimated the total of a calculation written on a board. Group 1 estimated the product of 8 x 7 x 6 x 5 x 4 x 3 x 2 x l while Group 2 estimated the product of 1 x 2 x 3 x 4 x 5 x 6 x 7 x 8. The earlier numbers in the sequence acted as anchors for the expectations, and Group 1 estimated a median of 2,250 and Group 2 estimated a median of 512. (The correct answer is 40,320 – I had to check it to believe it!)
Cognitive dissonance – if people cannot change their behaviour, then they change their beliefs about that behaviour, so there is less mis-match between belief and behaviour. The best example of this is the young smoker who feels unable to give up. They know that smoking causes cancer, but choose to believe it doesn’t do so in people as young and fit as them, who will, anyway, give up in the future. They know it makes breath and clothes smell, so they keep to the social circle of other smokers and choose to believe that ‘smokers have more fun.’
Festinger et al. (1956) empirically tested this theory by conducting a covert participant observation study of a cult that believed the world would end at midnight on 21 December 1954. They alone would survive and start a new civilization, and for this, a spacecraft was coming to pick them up earlier on that night. When the spaceship did not come and the world continued, then the cult members believed the world had been saved through their prayers.