Confirmation Bias Is the Mother Bias
Nobody googles "reasons I am wrong about this". We google "reasons I am right about this", and we smugly nod at how wise we are.
There are dozens of named cognitive biases, and this is the start of a series on them - I find them fascinating and I'm using this series as a means to learn about some psychology 101, and share what I've learned. One such bias sits under the others and is something of a basis for others. Confirmation bias is the tendency to seek, notice, remember and believe evidence that supports what we already think, and to ignore, discount or forget evidence that doesn't. It's how your uncle can keep bringing up the same political arguments at Thanksgiving dinner, forgetting what you've already said that clearly counters his argument - he won't remember, because the information you provide is irrelevant to him.
Evidence
The comforting model of the mind is a courtroom: you gather evidence, weigh it impartially, and reach a verdict. The actual mind runs the process backwards most of the time. You start with the verdict (the belief you already hold, or want to hold) and then work as its defence lawyer, scouring for supporting evidence and cross-examining anything inconsistent into silence.
You search for confirming information. You read ambiguous evidence as supporting your side. You remember the hits and forget the misses. You let confirming evidence in over a doorstep and make disconfirming evidence climb a wall[1].
Setting the Stage
I want to put two scenes down here, because I am going to keep returning to them throughout the series (or try to, if I can bend the examples enough each time!). Every bias in this run is easier to see when it happens to the same people in the same rooms, so: meet the dinner table and the dashboard. They will be back!
Why intelligence hinders you
Being clever, educated or numerate does not protect you (despite what most of us tell ourselves), and can make things worse. A smarter person is a better lawyer: quicker at generating persuasive arguments for what they already believe, better at spotting the flaw in inconvenient evidence, more fluent at explaining away the anomaly. It's funny how we assume intelligent people are wiser. There's a reason wisdom and intelligence are separate ability scores in pathfinder and dungeons a dragons: intelligence is knowing that a tomato is a fruit, wisdom is not adding it to your cereal.
Studies find that greater numeracy and reasoning ability can increase polarisation on politically charged questions, because people put their cognitive horsepower to work for their existing side rather than for the truth[2]. The ignorant are overconfident, fine, we all knew that. The capable are better at rationalising, which is not a great thing for society.
The one habit that helps (apparently)
Willpower does nothing here. "Be less biased" is an instruction on the level of "be taller". Confirmation bias runs below awareness, so telling yourself to be fair-minded mostly produces a more confident lawyer. The thing that helps is structural: go and find the strongest version of the opposing case.
Before settling a belief that matters, hunt down the best argument against it, from the cleverest people who disagree, and try to state their case so well they would sign off on your summary. That is steelmanning, and it is the discipline behind how I try to write (and mostly fail at).
Ask "what would change my mind?" and get worried if the answer is "nothing". Watch for the flinch, too. That little urge to wave a piece of evidence away is strong and itself deeply bias - you're doing that with your "tribe", your "side" and your "argument" in mind, not actually fairly comparing evidence. Your defence lawyer has found something and would like it kept out of the record.
But, even with that...
It won't make you unbiased. Nothing does. But - it will insert a step where you point your intelligence towards something more than building walls.
You cannot trust your sense of having weighed the evidence fairly. It is a horrible fact to know about yourself and we're kind of stuck with that fact. What's worse, knowing this doesn't solve the problem. It's like recognizing you having a proclivity for a certain flavor of ice cream doesn't make you like all of them equally suddenly. But maybe it can help you try new flavors, if you take that next step. This metaphor is a bit stretched: what I mean is, try thinking of other perspectives thoroughly. Intellectual empathy; emotional, too.
Go and read the Nickerson review. It came out in 1998.
Common questions
If everyone has confirmation bias, isn't objective truth hopeless?
No. Individuals are badly biased, but well-designed systems can be much better than any individual in them. Science works, imperfectly, because it institutionalises disconfirmation: peer review, replication, adversarial scrutiny, predictions registered. The fix for a bias you cannot remove from a single head is to build processes where other people's confirmation biases point in opposite directions and cancel it out, somewhat.
How is confirmation bias different from just having strong opinions?
Having a strong opinion is fine and often justified. Confirmation bias is about how you maintain it: protecting it by filtering evidence instead of testing it. What gives it away is not the strength of the belief but how the belief behaves when challenged. A well-held strong opinion can name what would change it and engages with the best counterarguments. A bias-protected one waves disconfirming evidence off without looking at it and never seems to update about anything, ever.
Doesn't steelmanning the other side just make me vulnerable to bad ideas?
I think that a belief challenged is a belief trained, like lifting weights. It hurts, and is tiring, but it is the only way to improve. If you argument is so weak that it can't stand up to alternative view points, why are you so keen to hold onto it?
References
- Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2(2), 175–220. https://doi.org/10.1037/1089-2680.2.2.175
- Kahan, D. M., et al. (2017). Motivated numeracy and enlightened self-government. Behavioural Public Policy, 1(1), 54–86. https://doi.org/10.1017/bpp.2016.2
