Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Crowdsourcing the availability heuristic

31 Post author: kalla724 25 April 2012 06:20AM

There are goals which can be achieved only by personal exertion and hard work – finishing a university degree, learning a language, mastering a martial art… But there is also a plethora of smaller goals, where small differences in approach and resources can make a huge difference. I’m going to examine how one particular cognitive bias affects execution of small-to-midrange goals, why this bias cannot be realistically overcome on a personal level, and how it can be effectively short-circuited simply by involving other minds.

Do note:  the point I’m making may seem obvious; in my personal experience, and from observation, it is one of those things that are obvious once you know the answer (and one still needs occasional reminders). The solution to the problem is high-impact and available to practically everyone, but remains vastly underused.

The bias in question is availability heuristic. LW has a decent definition:

The availability heuristic judges the probability of events by the ease with which examples come to mind. Sometimes this heuristic serves us well, but the map is not the territory; the frequency with which concepts occur in your thoughts need not reflect the frequency with which they occur in reality. Undue salience, selective reporting, even subtle features of how the human brain stores and recalls memories can distort our perceptions about the probability of events. Because it's easier to recall words by their first letter, people judge words that begin with the letter r to be more frequent than words with r as their third letter, even though in fact, the latter is more frequent. Or selective reporting by the media of dramatic tragedies makes them seem more frequent than more threatening albeit mundane risks.

This topic has been talked about many times on LW, and there is a great deal of academic research as well [1], including seminal texts many here will be familiar with [2]. It's all interesting, but there are two critical points I want to pull up to the forefront.

I - availability heuristic is commonly treated as a simple perceptual bias. In actuality, it is also a choice bias - if you fail to perceive an option, you cannot choose to pursue it. If you do pursue an option, you are likely going to focus on attaining it with cognitively available resources, while missing much better resources that are sitting idle. Similarly, you may waste resources (sometimes to the point of simply giving up) while pursuing a sub-optimal but cognitively available path towards your goal – completely oblivious to much easier roads which are actually available to you.

II - availability heuristic is not really "curable." Sure, it’s good to be aware of it. But we all have limited information about the world. Even if we objectively write down all known relevant factors for some observation or decision, our sample is still going to be at least somewhat biased, and certainly very narrow. Outside our direct areas of expertise, the amount of information we can include into any decision is quite limited; and the number of items our working memory holds while the decision is made is limited yet further.

One obvious mitigation strategy is probably apparent: simply research your desired goal by consulting experts (or the Googlian Oracle). How did other people achieve the goal? What preparations did they undertake? What strategies are recommended?

Good idea, which comes with several limitations.

Much of the information out there is written by people who are unaware of the existence of availability heuristic. The recommendations, however refined, are still usually descriptions of single strategies which happened to work optimally for the person who wrote the article. They could work (and often will), but they may not be the optimal solution to your particular problem set. Furthermore, as I will illustrate soon, the most common strategies you will find are the ones most cognitively available to the most people; an Internet search will, in effect, potentiate the availability heuristic even further, hiding less obvious strategies even further.

But by far the most significant limitation comes from the fact that knowledge does not equal resources. Even if you research an optimal strategy, you may still be unaware of the full scope of resources that are available to you. To avoid abstraction, let’s take a specific example.

Example: crowdsourcing adventure opportunities

A friend of mine has an interesting strategy for increasing the overall awesomeness of her life. Every January 1st, she comes up with a general rule, which she then follows until the end of the year. The rule is modified by common sense (you don't follow it if it will get you into an extraordinarily dangerous situation, or if following it is otherwise prohibitively expensive), but other than that, it has to be followed.

This year, the rule is "when you think or hear of something that causes you to be afraid, go ahead and do it." She's afraid of heights, so the obvious "go skydiving" is on the list from the start. But then, someone mentions flying in an acrobatic aircraft. That gets added to the list. I'm sure you see the general principle.

Seems a bit cheesy, at first glance. But... within the first four months of this year, she went flying in the aforementioned acrobatic aircraft (and a helicopter), learned how to ski, and even rode a damn ostrich. All within four months. She went caving this past weekend. Hang-gliding is firmly scheduled in a few months. And there are other, less glamorous experiences as well, but it's quite a list.[3, 4]

Now, let's say I wanted to do one of those things – say, an acrobatic airplane ride. What comes to mind? I look up places that offer such rides. I would need to travel there (and given the distances involved, get a hotel room as well). Pay the fee. Get to spend about 15-20 minutes being flown around. It's a lot of effort, and the payoff doesn't seem really worth it. Hey, let’s see what other people did! Google, google… a bunch of testimonials about people having a great/awful time taking the aforementioned touristy rides (most significant finding: corkscrews often cause explosive nausea). So I give up.

How did my friend do it? She talks to people about it; asks them for ideas. And soon enough, someone says "yeah, I know a guy who owns an acrobatic plane, wanna ride with him?" And lo and behold, a free ride of much higher quality than touristy nonsense one pays for, plus it's very close to home.

My approach above is the cognitively available one. I'm proceeding towards the goal in accordance with the patterns I've followed previously, when achieving similar goals. I'm thinking about resources that are available to me, personally (my money, my time, etc.). I end up with a suboptimal plan.

Her approach is to crowdsource: throw the desire into the world, and see what others come up with. Many people, with many different resources, ideas, and further links to even more people out there.

Once I started thinking about this, I decided to test this concept. I tried throwing out the acrobatic ride idea to my friends - and lo and behold, a friend of a friend of a friend is going to be flying in this summer. In his acrobatic aircraft. And now, when he gets here, I'm likely to get an hour or so of riding time with him, for free. Just because I asked.

It's a somewhat silly example, of course, but I think it illustrates the point. This friend (of a friend)2 is a resource. I was unaware I had this resource, until I asked for it. Finding an acrobatic aircraft owner through personal connections is a strategy that would have never occurred to me (since my availability heuristic informs me that such people are exceedingly rare).

Does this still seem obvious? Many articles were written about the “breakthrough” design of the Apple headquarters – a building made to force people together, to produce conversations between workers in different areas, and interactions between people who think in very different ways. In MIT lore, legends are written about Building 20, a “magical incubator” that has produced an incredible amount of breakthrough technologies and world-class thinkers. One of the main reasons given for this productivity is that many disparate small groups of researchers in a wide range of areas were thrown together in a small space – where they had to interact and talk to each other.

In other words, there is a certain kind of “magic”… in places that force people to simply utilize each other’s cognitive resources, to seek out different ways of thinking, and to avoid falling into cognitively available approaches to the problems they are trying to solve. And the point I’m trying to make is that one doesn’t need to work in Building 20 – just to intentionally maximize the utilization of their own social network (and work on diversifying it as much as possible).

There are people who already utilize their social network to the utmost, and who expand it strategically, adding people just to enhance the diversity of available viewpoints. But I will take a chance, and state that most of us probably don’t. And as a result, we aren’t able to recognize all of the resources available to us, to optimally use those we do recognize, or to realize optimal strategies for approaching our goals. Chances are that most of us could improve the strategy and execution of any given midrange goal – simply by asking around.

_____________________________

(EDIT) Addendum: help-seeking, status, etc.

There is a bit of discussion in the comments regarding some important questions - how and when does seeking help affect status within the group, would seeking help on a regular basis cause people to become uninterested in helping, etc. These may become a basis for a different text in the near future.

But these questions miss an important point here. Sure, asking for help can be a part of the strategy I discuss above, in some cases. In most cases, however, you should not be seeking help. The point of the article here is to simply talk about your goal and your strategy with others. "Involving other minds" does not necessarily require them to take an active helper role in the achievement of your goal.

In the specific example given in the text, I didn't go around saying "hey, I'm looking for help in finding an acrobatic airplane ride." Instead, I would say something like "Riding an acrobatic plane seems like an interesting thing; I'm trying to look into finding an opportunity to do so in the near future." Thoughts, offers and the eventual connection grew organically from the discussions that followed. Sure, the pilot himself is going to be doing me a favor (which I'll eventually repay), but the people who made the connection for me were just having a conversation.

To illustrate further on an example that popped up in the comments: a CEO of a company that always asks for help in making decisions will rapidly lose status (and therefore become ineffective at his or her job). This much is true. But most effective CEOs will organize their companies so that people of varying backgrounds will have to talk (at some point or other) about current company projects and strategies. The CEO doesn't ask for help in making the decision: she requires that her underlings produce ideas and overviews, which then become a basis for making optimal decisions.

_____________________________

References:

 [1] A few recent examples: Hayibor, S., Wasieleski, D.M. (2009). "Effects of the use of availability" Journal of Business Ethics 84: 151–165. Also, Klinger, D., Kudryavtsev, A. (2010). "The availability heuristic and investors' reactions to company-specific events" The Journal of Behavioral Finance 11 (50-65).

[2] Tversky, A., Kahneman, D. (1973) "Availability: A heuristic for judging frequency and probability" Cognitive Psychology 5 (1): 207–233.

[3] She's blogging her progress through this year, and the whole thing is highly recommended; for sheer hilarity as much as for some very interesting insights.

[4] One could write an excellent text on goal-setting strategies around this example. The rule is simple and absolute ("never have any chips, cookies or other snacks available at home" is easier to follow and will in most cases lead to a greater weight loss than an intricate diet), and it is overarching (applies everywhere in life, not only to some particular times and places, making lawyering around the rule much more difficult). If you are going to set rules for yourself, this is the way to do it. But since I'm writing a loose set of texts on a completely different topic, this footnote will be all I have to say on that topic.

 

Comments (40)

Comment author: Mark_Eichenlaub 25 April 2012 07:19:46AM 13 points [-]

I like the idea. Perhaps we should start a periodic discussion thread where people post midrange goals and get feedback.

Comment author: komponisto 27 April 2012 01:49:51AM 8 points [-]

Whatever the merits of this post (and I'm willing to admit they may be considerable), the usage in the title doesn't make sense to me:

Hey, let’s see what other people did! Google, google… a bunch of testimonials about people having a great/awful time taking the aforementioned touristy rides... How did my friend do it? She talks to people about it; asks them for ideas....Her approach is to crowdsource

So you're using Google -- effectively the collective brain of the entire planet -- while she's talking to a few people she knows....and she's the one who's engaged in "crowdsourcing"??

I grant you that her approach may be the better one, but do you really think "crowdsourcing" is the appropriate term for this contrast?

Comment author: kalla724 27 April 2012 09:29:44PM *  2 points [-]

I see your point, but I'll argue that yes, crowdsourcing is the appropriate term.

Google may be the collective brain of the entire planet, but it will give you only those results you search for. The entire idea here is that you utilize things you can't possibly think of yourself - which includes "which terms should I put into the Google search."

Comment author: chaosmosis 30 April 2012 09:34:16PM 1 point [-]

In real life, you can only ask the people who you're already friends with. That means you'll probably share common biases.

Unless you asked strangers. That might be a good way to fix this.

Comment author: fburnaby 30 April 2012 08:25:56PM *  0 points [-]

Yes. The art of Googling can be pretty difficult, and a few brains are still smarter (though less broadly knowledgeable, perhaps) than Google, at this point in time.

Comment author: shminux 27 April 2012 10:16:12PM 1 point [-]

A standard name for this is networking: you explain your need to people IRL, or in your blog or wherever, and hope for useful feedback. The replies you get will be customized to you and your relationship with those reading/listening. No one is likely to offer a free plane ride to a total stranger.

We all know that some of the best opportunities are never advertised publicly, be it a plane ride with a friend, a manager's special at a store, or a covetous job opening. There are many more opportunities for couch surfing than posted on couchsurfing.org, etc.

Comment author: thomblake 25 April 2012 07:20:30PM 7 points [-]

The "failure to look up options" version of this is a major source of forehead-slappage for me.

"Why is [unclogging this sink / gluing together this broken vase / reformatting ths hard drive] taking so long? There has got to be a better way! Wait, why do I think I know how to do this in the first place? *Check Google*... Dammit, I was doing it wrong!"

Comment author: TheOtherDave 25 April 2012 09:07:10PM 8 points [-]

"Wait, why do I think I know how to do this in the first place? "

This is going on my list of questions to ask about any new project.

Comment author: thomblake 25 April 2012 10:09:14PM 2 points [-]
Comment author: fiddlemath 25 April 2012 10:34:09PM 3 points [-]
Comment author: kalla724 26 April 2012 05:12:58AM 0 points [-]

True enough. Anything can be overdone.

Comment author: thomblake 01 May 2012 06:43:26PM *  0 points [-]

It is really just a form of the fundamental question of rationality, "What do you believe, and why do you believe it?"

Comment author: TheOtherDave 01 May 2012 06:57:36PM 0 points [-]

Sure, that's true.

And for someone who finds that remembering the fundamental question brings to their mind all the appropriate context-specific questions that follow from it, presumably all that person needs to do is keep the fundamental question in mind. (Similar things are true of "What should I do next, and what do I expect will happen if I do?")

Sadly, I am not yet that enlightened, so I find that keeping some of the context-specific questions explicitly in mind can help.

Comment author: thomblake 01 May 2012 07:02:02PM 0 points [-]

Indeed, I can't even keep that version in mind, despite it being helpful every time.

Nor, tangentially, can I remember to duck my head when getting into a car, ever since getting my Saturn Vue. Whack "Oh yeah, that happens every time."

Fortunately, I do appreciate physical comedy.

Comment author: arundelo 26 April 2012 12:30:30AM 3 points [-]

less glamorous experiences

"My initial excitement at the use of the serial comma was immediately dampened by the lack of noun after the hyphenated adjectival phrase."

I like this blog so far.

Comment author: John_Maxwell_IV 25 April 2012 06:26:40PM *  3 points [-]

It seems like this strategy should work better for highly unusual goals, like the ones you give as examples, that make for more interesting conversation.

I agree we should ask one another for help and advice more frequently; it's too bad that doing so is a status-lowering act.

Comment author: kalla724 25 April 2012 10:12:13PM 3 points [-]

It's not, not necessarily. There isn't as much research on help-seeking as there should be, but there are some interesting observations.

I'm failing to find the references right now, so take with several grains of salt, but this is what I recall: asking for assistance does not lower status, and might even enhance it; while asking for complete solution is indeed status-lowering. I.e. if you ask for hints or general help in solving a problem, that's ok, but if you ask for someone to give you an answer directly, that isn't.

But all of that is a bit beside the point. In the abovementioned approach, you aren't really "asking for help." You are just talking with people, telling them what you wish to achieve, and asking for their thoughts. They can choose to jump in and offer help if they want (which can be, and most often is, a happiness-enhancing action for themselves as well as for you).

Comment author: thomblake 25 April 2012 10:57:15PM 1 point [-]

For status amongst smarty-folks, asking for help definitely puts a ceiling on one's status. You are forever denied the realm of the self-sufficient, infallible genius.

Comment author: paper-machine 26 April 2012 06:38:30PM 1 point [-]

Is that really a title that one should aspire to?

Comment author: thomblake 26 April 2012 06:45:26PM 1 point [-]

Well, I don't act as though it is.

Comment author: John_Maxwell_IV 26 April 2012 12:06:29AM *  1 point [-]

I would be careful extrapolating results from one study, which only has one trial scenario. See these comments of mine:

http://lesswrong.com/lw/bs0/knowledge_value_knowledge_quality_domain/6d88
http://lesswrong.com/lw/bs0/knowledge_value_knowledge_quality_domain/6d94

What you describe would be a status-lowering act for a CEO if she was doing it with her employees (i.e. such a CEO might be praised for her egalitarian leadership style), but it's only a little status-lowering if you do it with friends you are on equal terms with. The more advice/help you ask for, the more you lower your status ceiling. Thought experiment: imagine a CEO who asked her employees for input on almost every decision she made.

This is unfortunate (seems possible that it's an efficient use of resources to get input from friends on all of your major life decisions and problems; I know I think better about other people's problems than my own). It's not obviously unsolvable though.

It probably matters a lot the sort of person you are talking to. I suspect I have a strong tendency towards assigning status based on intelligence only, so hearing about someone's problems doesn't cause me to assign them lower status. (I've also taken a number of acting classes, which partially left me with the alief that status is a big game that doesn't really matter.) I assumed this stuff was true for other people as well, tried to get them to debug/improve me, and found out the hard way that they would permanently lower my status for this.

Comment author: aelephant 01 May 2012 03:07:23AM 1 point [-]

One thing that occurred to me reading this was that a somewhat effective (not 100% effective) way of "guarding" one's status would be to solicit for help / advice in private rather than in pubic. Obviously you can't guarantee that people will not disclose things, but if you are in a position like the CEO in your example, you could take measures to encourage your employees not to share the contents of confidential discussions and so forth.

Of course, this has costs and benefits as well: it is much easier to make one post on a public blog than it is to private message everyone on your Facebook. If you're concerned about maintaining your status, it might be worth considering. For example, for someone with high status and a relatively "forgivable" question (difficult, not obvious) just posting it in public is probably better than going through the hassle to make the inquiry private; for someone with low status asking a less "forgivable" question ("He really should know better, did this guy not listen to what the prof said in lecture?"), the extra effort might well be worth it.

Comment author: adamisom 25 April 2012 07:33:47PM 1 point [-]

I'm not sure it's a status-lowering act. I know it intuitively seem so--I think if you're only ever asking, and not ever contributing, then yeah. But not if you contribute a lot too.

Comment author: John_Maxwell_IV 26 April 2012 12:12:41AM 2 points [-]

Giving advice is a status-raising act, so if you give and receive advice in equal quantities they cancel each other out.

Comment author: thomblake 26 April 2012 03:45:20PM 1 point [-]

so

That doesn't follow. Do they really raise and lower status in equal proportions? Regardless of initial status?

Comment author: John_Maxwell_IV 27 April 2012 06:19:11AM 1 point [-]

As long as we're on the topic of status, I'd be interested in using the grandparent to this comment as a case study. At first I took your response as an indicator that I should have phrased the grandparent less confidently (i.e. lower status), but then I realized that you might not have pointed out what you did if my comment wasn't so direct, and the conversation overall is probably nice and crisp to read. But then it occurred to me that if you hadn't come along, people might have become too confident in the grandparent. Thoughts?

Comment author: thomblake 01 May 2012 06:42:09PM *  0 points [-]

Thoughts?

No, no thoughts.

Comment author: John_Maxwell_IV 26 April 2012 05:20:51PM 0 points [-]

Fair enough.

Comment author: dlthomas 25 April 2012 08:04:22PM 1 point [-]

I'm not sure it's a status lowering act. I am sure it usually feels like a status lowering act, and so people avoid it.

Comment author: fburnaby 30 April 2012 08:30:24PM 0 points [-]

It definitely feels status-lowering to me when I ask for help. Consequently, I very rarely ask for help. However, I've noticed that my friends who do ask for help or do other things that feel status-lowering to me (especially "over-sharing" their feelings) also have more friends and more active social lives than I do.

Comment author: dlthomas 30 April 2012 09:08:31PM 0 points [-]

That "those activities are not actually status lowering" is certainly a potential explanation of what you observe, although it raises the question of why you (and so many others, it seems) interpret them to be. There are doubtless other explanations, however; off the top of my head: "having higher status for other reasons, the friends in question feel more able to engage in status lowering acts."

Comment author: fburnaby 30 April 2012 09:37:14PM *  4 points [-]

I'm just speculating at random now, but the idea has popped into my head so I'll share.

We're adapted to function in small tribes where status may be very absolute and worth guarding on the one hand, and cooperation/helping each-other necessary very frequently. Our modern situation isn't quite the same - I'm completely self-sufficient in the sense that I can participate in formal and impersonal business activities and then purchase anything I need. Most of my friends are the same - if we're out, we pay our own tabs; if we're having a bad day, we try not to spread our contagious bad moods to each-other.

But I've recently been reading Robert Wright's book "Non-zero". He suggests that trading favours with people is a central part of human bonding. We may need to have opportunities to get a feel for each-others' characters by exchanging small favours, before we start trusting each-other with bigger things (Is this person a defector? I'll test that out by exchanging a fairly trivial favour. If they don't defect, I can up the ante. Etc). If that's true, then we're not getting many opportunities to show each-other that we're co-operators, not defectors.

Of course, there used to be two ways of being a defector: 1) ruthlessly cheating for gain, or 2) being an inadequate tribe-member who can't carry their weight. In that sort of situation, requiring help too often would look bad in the same way that someone with bad credit would look to a lender - not someone to do business with. Just as private companies have "optimal debt ratios", perhaps humans do, too. If you're too needy you start to look like bad credit, but if you aren't needy enough, you never get an opportunity to up your credit rating. Perhaps the credit-rating -> status analogy has something for it. And perhaps relative loners like me are too more tuned to the "avoid being perceived as an inadequate tribe-member" logic than is appropriate in our wealthy modern world.

Comment author: John_Maxwell_IV 01 May 2012 01:14:32AM 0 points [-]

Or you could be programmed to act that way because you're a loner. You don't have much in the way of connections or credibility, so asking for favors out of the blue is a recipe for failure.

Comment author: fburnaby 30 April 2012 09:42:14PM *  0 points [-]

I agree that being "approachable" might play in the dynamic, too. Needing help may attract others who can thereby raise their own status by helping you.

Comment author: John_Maxwell_IV 01 May 2012 01:18:28AM 3 points [-]

I remember reading that getting someone to help you is a better way to make friends with them than helping them. (Although this may be due to consistency effects.)

Comment author: Ezekiel 25 April 2012 11:30:45AM 2 points [-]

I like it, and it raises an interesting question: How would one design a properly blinded experiment to check how effective a strategy is? You sort of have to tell the participants what they're doing, and with strategies like this one it'd be blatantly obvious what the goal was.

Until we can actually do that, though, this looks like it should work. I will bear it in mind for when I have some untreated short-to-midterm goals.

Comment author: asparisi 25 April 2012 11:00:03PM 0 points [-]

Well, you couldn't test against nothing without running into those issues, but you could compare strategies against each other, which can be more robust anyway.

So, you tell group A to try strategy 1 for a type of problem, and you tell group B to try strategy 2. You measure as best as possible the relationship between commitment to the strategy, effectiveness, and akrasia and later you compare and see who accomplished the most. You could tell each group that "(1 or 2) is the best method we've found for accomplishing (X/Y/Z) and we'd like you to track your results with it." which would act as a fairly effective blind as long as group A and B don't know about each other.

If you really want to test against a baseline, you could first try to discern what the "common" strategy is for most people and make that your control.

Comment author: Alicorn 25 April 2012 04:36:40PM 0 points [-]

Can anyone within reading distance of this post teach me to fly ultralight aircraft? I think I've flung this interest into the world before but I haven't gotten any results yet.

Comment author: nekomata 26 April 2012 02:05:43PM 0 points [-]

This is a strategy that I sometimes implement and I think it should be used more often, be it only because it allows friends of friends to meet or get to know one another better and that makes nicer social networks.

However, I wonder if it can be taken too far. If everybody routinely makes this sort of demands, won't people stop to answer them ? Is this technique successful partly because it is not widely used ?

Comment author: kalla724 26 April 2012 08:30:52PM 0 points [-]

I may have to edit the text for clarification. In fact, I'm going to do so right now.