Cognitive bias

[ciphergoth]

Roughly, "cognitive bias" is the empirical study of systematic, irrational biases that we all show to some extent or other in the way we think about the world. These biases can be demonstrated in controlled experimental settings, where we can largely rule out rational explanations for the behaviours seen.

One example is anchoring: asked whether they thought an unknown quantity was more or less than a number produced in front of them using a roulette wheel, subsequent guesses at what the number was were irrationally close to the number the roulette wheel produced. Their guesses had been "anchored" on the number they'd previously been given, even though they knew it was a random number.

I'm curious to know whether this is something people think about much, hence this (fairly imperfect) poll. Where I say "people" below, I mean people demographically roughly like you.

For those not familiar with the convention "snowflake" means "none of these answers fit what I'd like to say, so I'll comment below and explain". thehalibutkid, we await your comment :-)

[Poll #1327489]

I'm thinking about this because I've just finished reading "Mistakes Were Made (But Not by Me)", and I'm about to start on "Predictably Irrational", both of which are pop science books in this field. See also Wikipedia's list of cognitive biases.

Please do comment with any thoughts the poll doesn't cover, of course!

Comments

[wildeabandon.livejournal.com]

Was "Mistakes were made (But Not by Me)" any good? It's been on my list of books I should get round to reading for a while now.

[ciphergoth.livejournal.com]

Not that fantastic, a bit repetitive. I'll lend it to you when we do lunch :-)

[wildeabandon.livejournal.com]

Heh, thank yee. And yes, lunch is a good plan.

What's your diary like? I could do Monday-Wednesday, but after that I'm away or in all day meetings until the 26th.

[djm4]

For the first answer, I ticked 'snowflake' because, while I think I've got a pretty good understanding of many of the issues of cognitive bias, and it's an issue I've been wrestling with for at least twenty years, I'm not familiar with the term itself or much of the jargon cited in the Wikipedia article.

So, in a sense, I've never heard of it, while still having a lot of layperson's knowledge about it.

On the last two, I ticked 'always', but only because you didn't have a 'usually' option, and 'sometimes' seemed to understate how often I use it.

(Edited for typo)

[ciphergoth.livejournal.com]

Heh. I answered the last two differently because I don't think I'm nearly as rigorous in looking for bias in my own decision making as I am about other people's, and that's something I'd like to fix but I'm not certain of the best way to do so...

[wildeabandon.livejournal.com]

Hmm. I (think I'm) more rigorous at looking for my own bias, and suspect that comes from having views which are out of sync with a lot of the people I mix with. Clearly you need to start hanging out with more Xian Tories ;>

[ciphergoth.livejournal.com]

I'm not sure that's the right approach :-) For one thing, I've come to realise that one of the attractions of talking about atheism for me is that I could hardly be more confident about what I think or my ability to defend it, and it's likely I should spend more time talking about things where I feel on a less sure footing (which is practically everything else).

[palmer1984.livejournal.com]

Yes, I like talking about atheism for that reason.

[palmer1984.livejournal.com]

Well, actually not just that; also I'm much more certain about atheism than I am about my views on most other things!

[ciphergoth.livejournal.com]

I had meant my answer to imply that too!

[palmer1984.livejournal.com]

*looks again* You did indeed. I misread you!

[drdoug.livejournal.com]

By default I tend to think that I'm better than average about being rigorous about looking for my own biases (and correcting for them), and also better than average at looking for other people's. But I worry sometimes that it's just my fond imagination.

[djm4]

I tend to veer the other way, because I feel that I'm often more likely to be able to identify some of the biases in my own decision making. This means that while I can consider cognitive bias in other people's decisions, I may not have the information to do so productively.

[olethros.livejournal.com]

I'm sort of interpreting my behaviour into the questions above, but I reckon this applies.

I try to be as aware as possible of the margin for error in my understanding of others' actions (and reactions, when examining my own actions.) It really annoys people sometimes, because I've gotten awfully good at finding reasonable (not rational, reasonable) explanations for the behaviour of all sorts of annoying people. :-)

[bohemiancoast.livejournal.com]

Anyone who believes that they are as rigorous in seeking out their own bias as other people's should consider whether they know anyone else who can do that. It's very hard. I suppose the sovereign cure is to discuss things with other people before reaching conclusions, but still. I'm a very intuitive decision maker (that means I can rarely articulate the decision making process), and although that is not inherently a more biased approach than structured decision making, it's risky.

[faerierhona.livejournal.com]

I snowflaked on the last two as I don't consciously use it, but I am sure the little knowledge I have on it affects how I view others decisions

[mooism.livejournal.com]

It's something that interests me but scares me (in that I sometimes avoid reading about it to avoid the pain of the fear; I don't lack interesting things to read so there is always something else I can read instead).

This sort of thing should be taught in school.

[countess-sophia.livejournal.com]

I understand and recognise the concept, and have seen it in action frequently, I'd just never heard this name for it before.

Soph x

[palmer1984.livejournal.com]

Me too.

[skibbley.livejournal.com]

I'm interested mostly right now in how bias occurs out of conciousness from exposure to the media etc. and ways of counteracting that. Applications include countering prejudice against women, gay people, black people etc.

I'm also feeling increasingly shaky and interested in how people can not accurately know their own motivations and why they made past decisions.

[skibbley.livejournal.com]

Oh, and I think one of the biggest things happening is people becoming more and more fearful while becoming more and more safe.

[ciphergoth.livejournal.com]

Have you read "Blink"? The discussion of the implicit association test and the ways it reveals on how to reduce unconcious bias is pretty interesting.

One happy conclusion is that having a black President is likely to substantially reduce racism simply because it will give people a new association to go with a black face.

[skibbley.livejournal.com]

Yes, and I hadn't made the connection to a black President.

"Risk: The Science and Politics of Fear" by Dan Gardner is pretty good although I am intrigued as to why Gardner and Bruce Schneier don't reference each other.

[deliberateblank.livejournal.com]

Because they're both trying to sell books discussing exactly the same thing, and most people only have room on the shelf for one?

[jhg.livejournal.com]

Interesting - that reminds me of something RTD said regarding the visibility of same-sex relationships - i.e. on telly etc. - as a way of combating homophobia.

In short he said that people have a picture in their heads of a man with a woman. 2 men together breaks that picture, and so people are confused/scared/hateful etc.

The only way to break that is by visibility - showing happy same-sex relationships etc.

[bohemiancoast.livejournal.com]

Recognising and dealing with cognitive bias, in ourselves, our colleagues, and our stakeholders, is a critical part of policy development. We don't always do it as well as we could.

This comment unselfconsciously girly, no content

[valkyriekaren.livejournal.com]

Pretty passionflower icon!

[simont]

I'm not entirely sure you can get away with saying something is unselfconsciously girly, when you've clearly demonstrated consciousness of it by saying so!

(Unless that was a deliberate cognitive bias for comic effect? :-)

Re: This comment unselfconsciously girly, no content

[bohemiancoast.livejournal.com]

Thank you! It's a photo I took of a passionflower in my back garden a few years ago. It's only later that I discovered how uncommon photos with both the flower and fruit are.

[jhg.livejournal.com]

I've read various articles regarding this in the course of my Philosophy degree, and I certainly do find it interesting... but not to the extent of actively reading about it now.

[ludy]

i snowflaked on the question about other people's decisons because of the whole autistic finding it hard to get into other people's heads thing.
As far as my own decisions go how much thought i give to basis depends largely on my spoon levels and how much mental time and space i have spare. Which prolly means i'm least likely to consider biases at times when i'm most vunerable to them ...

[ludy]

(hm ... i've just realised going back to an unpaid account means i can't edit comments - bum!). I'm not familiar with all of the jargon about this bit i did study some of thease cognitive bias stuff in my peace studies degree - particularly as it relates to conflict resolution

[http://users.livejournal.com/_lj_sucks_/]

It occurs to me that this is why sales pitches often take the form "You might expect a luxury car like this to cost $40,000, or even $50,000, but it's just $29,999!"

[zwol.livejournal.com]

I've done research on things that could be interpreted as very limited subcases of cognitive bias, but I don't think they are best interpreted that way, 'cos they don't generalize well when you do. So my knowledge of cognitive bias itself is fairly fuzzy.

[kissmedeadly.livejournal.com]

Snowflake: I'm trained as a counselling psychologist, so while I pretty much did the standard reading for cognitive stuff in my undergrad degree, theres certain very specialised areas of it that you tend to use far more in clinical work. It usually came up if I was working with phobias, anxiety and/or depression, and fits in nicely with CBT oriented work.
And before your brain heads to the most used acronym for your particular interests, no, NOT that CBT, nor does it have anything to do with motorbikes, its cognitive behavioural therapy.

[henry_the_cow]

The poll is a little self-selecting, of course, because people who are interested are more likely to respond.

I snowflaked on a couple of questions because the term "cognitive bias" as described in the Wikipedia article seems to cover a wide range of different phenomena. Some of these I consider - e.g. probabilities, cause-and-effect, wanting to look good, hindsight - although I haven't grouped them together or thought of them as "cognitive bias" per se. Others I don't - e.g. anchoring.

I think I'm poor at looking for such mistakes in my own reasoning and more likely to spot them in other people's thinking. I'm generally better at thinking about things than about people (including myself).

[anarquistador.livejournal.com]

Sorry to stry off-topic, but any theories as to where "snowflake" comes from?

[bohemiancoast.livejournal.com]

It is short for 'I am a unique and precious snowflake and my answer isn't one of your options', which in turn is riffing on the alleged individuality of snowflakes. I suspect some internet wag came up with it first. I think my friends have been using it for a couple of years.

[pavlos.livejournal.com]

Snowflake: I don't recall enough instances of examining my own decisions to answer the last question. I don't think I'm exceptionally self-uncritical, I just deal with other people's decisions that much more often.