|
– Francis Bacon
(Attentional) Confirmation Bias A pervasive cognitive failure resulting in thinking about positive rather than negative information pertinent to one’s current beliefs, and involves people’s tendency to seek, attend to, or to remember information which is consistent with and can therefore support – but not counter – a belief, preconception, or expectation.
People are more likely to accept supporting evidence at face value while subjecting evidence contrary to a particular, preferred conclusion to more critical evaluation. This can lead to the unpleasant result that the more deeply an actor holds given beliefs, after confronting evidence which disagrees with those beliefs, the stronger (polarized) they are liable to become in their initial beliefs† – counter-evidence may therefore be easily resisted.
The following “card problem” gives a simple illustration of this positivity bias: there are four cards marked “A”, “B”, “2”, and “3”. Each card has a letter on one side and a number on the other. The task is to determine whether all cards with a vowel on one side have an even number on the other – two cards may be turned over in an effort to decide the problem. Whilst people unfamiliar with the puzzle almost always choose “A” and “2”, “3” would be a better choice. Why? Selecting a “2” can only bring to light that some cards with a vowel are paired with an even number: one is little better off than before turning over the “2” card. Whereas finding a vowel on the back of the “3” card would deal a death blow to the hypothesis at issue: whether all cards with a vowel on one side have an even number on the other side. (see: falsification) The tendency to selectively gather information which is consistent with prior expectations has been illustrated by various research such as studies in which subjects conducted interviews to ascertain the personality traits of interviewees. Half of the subjects were asked to determine whether interviewees were extraverted (outgoing, sociable, gregarious, etc.), while the remainder were invited to find out whether interviewees were introverted (shy, reserved, timid, etc) – all subjects used a list of experimenter-supplied questions for assessing extraversion and introversion from which they selected questions to ask interviewees.
Subjects looking for extraversion preferentially selected questions slanted towards a presupposition of extraversion, for example, “What do you do if you wanted to liven things up at a party?” While subjects on the trail of introversion chose questions assuming introversion, for example, “What factors make it really hard for you to open up to people”. Thus, not unsurprisingly, the answers subjects were given by interviewees – in light of the questions posed – were such as to confirm the subjects’ initial beliefs. They essentially ‘found’ the personality traits they had been seeking in spite of, rather than because of, the interviewees’ actual personality traits: on the basis of the questions the subjects’ themselves decided to ask!
Such findings should be a salutary lesson for those who in their professional lives lean towards the extensive use of personal interviews for assessment purposes, such as recruitment and staffing procurers.
See Carroll for an extended discussion of confirmation bias.
† this is called ‘congruence’ – people’s tendency to retain an initial – and oft-times favoured – hypothesis. (see also: Belief perseverance, Post hoc reasoning, self-fulfilling prophesy, myside bias, Subjective Validation, belief bias, cognitive bias, Selective exposure, Forer (Barnum) effect, cognitive illusion, Illusion of validity, file drawer problem)
Back to: Glossary A-Z
Glossary of selected Judgement & Decision-making, Belief-related, and other Psychology terms A-Z » » Return to belief, judgement, and clear thinking » Labels:
Confirmation bias, attentional confirmation bias, Confirmation bias – attentional, bias – confirmation, bias – confirmation (attentional), definition: ‘attentional confirmation bias’, definition: ‘confirmation bias’, biased thinking, congruence |