Confirmation Bias

Wikipedia defines ‘confirmation bias’ as a tendency to favour that which confirms your existing belief. (“Confirmation bias, also called myside bias, is the tendency to search for, interpret, or recall information in a way that confirms one’s beliefs or hypotheses” Mar 2015).[1] As such it defines a psychological phenomenon or perhaps we might even say a social-psychological phenomenon. For example, we may collect and interpret data in a way that favours confirmation of our hypothesis, that is to say, we may display a tendency to do this when we ought to be more critical. Or again, we may interpret texts, ‘problem’ passages, so that our doctrine is not challenged. In short, we have a bias to see only what confirms our belief in something.

When arguing with someone, to observe that they are displaying confirmation bias can be helpful; equally it can be a rhetorical strategy to cast doubt on an opponent’s argument without engaging their points. In using such a strategy, we could be displaying a reverse confirmation bias ourselves because we find what someone is saying unpalatable. Thus, someone may be interpreting data consistently in a way we find unacceptable, perhaps overlooking or smoothing over recalcitrant details, and we might say that they are biased in their handling of the data. Equally, however, it may be that it is us who are biased; it is just that we are blinded by our good opinion of ourselves.

These opening observations should cause us to think twice about charging someone with confirmation bias and rejecting out of hand such a charge levelled at us. What we should be careful to maintain in argument, however, is the distinction between logic and psychology.

It may seem intuitively obvious that truth-value is a quality of propositions, and that being true or false is established through logically sound reasoning. The fact that we might be biased has no causal bearing on whether what we propose is true or false. The oldest intuition in this area is that things are true and false independently of us. Gottleb Frege, the father of modern logic, separated logic from psychology; Frege’s main criticism of ‘psychological logic’ was that it conflated ‘true’ and ‘being-taken-to-be-true’.[2] We could actually ask whether a tendency to make the accusation that someone is displaying confirmation bias is itself a bias in favour of a psychological explanation for reasoning rather than a logical evaluation of reasoning. Those with debating skills may display this bias because it is effective, but it seldom addresses the issue of what is true and false.

What is the correct response to an accusation of confirmation bias and should we ourselves make such an accusation? You should certainly reflect privately on the accusation but defending yourself against the accusation is inadvisable because this puts you in the foreground of debate rather than the argument. The Christian principle is generally not to defend yourself. This means that the one who accuses you of confirmation bias will see their charge unchallenged. This means that onlookers listening to the argument could be deflected from the logical merits of the argument you are making because you have not defended yourself against a charge of confirmation bias. They should instead see that a charge of confirmation bias may be a rhetorical strategy to avoid engagement with the logic of an argument.

The phenomenon of confirmation bias is relevant to evidential reasoning in which there is a belief or hypothesis that is confirmed (or not) by bringing evidence to bear on that belief or hypothesis. Clearly, therefore, whether a person is displaying confirmation bias is a matter that is also either true or false for which evidence can be put forward, and indeed it might be observed by some wag that the evaluation of that evidence for confirmation bias could itself show confirmation bias, especially if it is just anecdotal. This is the beginning of a hopeless regress and it shows that when arguing turns from logic to psychology, it is time to walk away.

There is a further dimension. It is one thing to have particular beliefs to which you are attached and for which you may be biased, but religious thought is a systematic worldview and beliefs cohere together in a closely connected network. The danger of bias on a particular point is hidden by the network of supporting beliefs. This makes interpretation rather than evidential reasoning the issue. Your interpretation of all the data may be biased in one way or another. One just has to think of how false doctrines are supported by an interpretation of all the relevant data, including ‘problem’ passages, to see that this is the case. Worldviews clash; a systematic interpretation of all the data is met by an opposite configuration of all that data. So, while there will still be argument, an accusation of confirmation bias in your evidential reasoning for a particular belief misses the point because your whole position on all the data is what is at issue.[3]

Is defending yourself against an accusation of confirmation bias productive? We have said the accusation is an effective strategy in a debate with an opponent. Against such an accusation, in a debate, the best counter-strategy is to distinguish logic from psychology and stay with the logic of the argument(s) rather than defend yourself. It is different in the co-operative environment of the research seminar, the ecclesial bible study, or the lab, where there should be open-mindedness and this should mean that the confirmation bias of any one individual is countered by the co-operative spirit of the group.

To conclude: someone might ask whether confirmation bias is a logical topic. The answer is that it helps define what logic is not – it is not concerned with psychology. Logic is concerned with reasoning constructed in language and hence it has an objective quality and for this insight we are indebted to Frege.