The wisdom of crowds principle is that if you gather a group of diverse people who are not idiots, then ask them to estimate something, and average their answers, their errors will be different and tend to cancel out. The result is that the averaged answer will be more accurate than the answer of a randomly chosen group member.
The SciAm article describes how you can simply make two estimates for a problem, and get some of the same advantages...if you use a process for your second estimate that stacks the deck in favor of mitigating your errors:So you can actually apply some of this recent research without gathering a crowd.
“First, assume that your first estimate is off the mark. Second, think about a few reasons why that could be. Which assumptions and considerations could have been wrong? Third, what do these new considerations imply?... Fourth, based on this new perspective, make a second, alternative estimate.” When the participants used the more involved method, the average was significantly more accurate than the first estimate. The “crowd within” achieved about half the accuracy gains that would have been achieved by averaging with a second person.
I just realized that here I'm using the popular misconception for schizophrenia in my title...referring to split personalities, which are extremely rare, while schizophrenia itself is pretty common. Wikipedia says it's commonly described as affecting 1% of people but that a review of multiple studies came up with a smaller .55% figure. We're talking about something like 1 in 200 people having schizophrenia at some point in their lives. The true workings of schizophrenia, which happens when your sense perceptions get messed up, are pretty interesting; a good discussion of it is found in The Throwing Madonna: Essays on the Brain, which I read recently.
So, I can go through a complicated process of creating a competing viewpoint somewhat isolated from my own biases on a subject just to get to a conclusion roughly half as accurate as I could reach just by talking to one other person, say you or Lisa?
ReplyDeleteWow. This is a very good argument for having friends you can talk to. But if I'm ever alone on a desert island or I need an opinion on something too embarassing to share with others, like porn or a colonoscopy, I might try it.
You have to sign your posts, Anonymous Doug. Anyway, you're being pessimistic and giving bad examples -- how difficult is it to estimate an opinion about porn or colonoscopies? This isn't about viewpoints, it's about estimates. Like, say you have to give a schedule for something...like a freelance writing project.
ReplyDeleteIt's me again! Doesn't the quote from the article say that test individuals averaged half the gains in accuracy provided by sharing the estimation with one other person? That implies to me that I'm better off cultivating a friendship with someone smart like you or Lisa or even floating an idea before an anonymous collection of forum drones to solicit their opinions on an idea than I am mastering this skill. I just read an article that says Americans isolate themselves more than any other culture and this study just seems to reinforce the myth of self-reliance and avoiding contact with other people. (and for some reason I can't use my normal ID from my iPod) Doug
ReplyDelete