Anecdotes vs statistics: Who do you believe?

By William Orem

Posted September 9, 2024

Media Player Error
Update your browser or Flash plugin

Transcript

D: On today's moment of science, we're going to talk about aspirin. Aspirin is a "pain-killer." They're called pain-killers because they make pain go away.

Y: That's not true. Aspirin doesn't work.

D: Whaddya mean, Yaël?

Y: Sorry. But you can't go telling people that aspirin works when I have personal experience that proves otherwise.

D: You do?

Y: Sure. My cousin Byron gets headaches all the time. He says aspirin never works for him.

D: So you assume it won't work for you either?

Y: Of course not! Listen, Byron is no liar.

D: Suppose I told you that ninety percent of the people who take a certain drug report that it works like a charm. Would you try that drug?

Y: (thinking): Well, I suppose so. Ninety percent seems pretty high.

D: Well, suppose I said that ninety percent of the people who use aspirin say it works?

Y: I told you, my cousin Byron...

D: Aha! So you'd be willing to ignore lots of compelling statistical evidence in favor of one anecdotal account.

Y: (puzzled): Well...I guess I did just say that...

D: You're not alone. Social psychologists are familiar with this effect, sometimes called "the man who" effect. I say: ninety percent of the people who try this drug receive benefit from it. You say "I know a man who tried it and nothing happened!" Somehow the face-to-face testimony of one person seems more convincing than any amount of statistics.

Y: Hmmm...I guess that's not very rational.

D: Nope. Because those statistics represent lots of individuals just like Byron.

Y: Well, nobody's quite like Byron. He makes a living pounding in nails with his forehead.

D: Well, that explains it...!

Round, white asprin pills all in a grid pattern against a light blue background

Social psychologists are familiar with something called "the man who" effect. (Daniel Foster / flickr)

Countless people take aspirin every day as a pain reliever, so it must be pretty effective. But let's say you have a cousin who swears aspirin doesn't work for him, so you think it might not work for you either. Do you still take it?

If ninety percent of the people who take a certain drug report that it works like a charm, wouldn't you try that drug? Ninety percent seems pretty high. As soon as you hear that drug is aspirin though, you hesitate. Your cousin is no liar, so it might make you stop and think.

Do you ignore lots of compelling statistical evidence in favor of one anecdotal account?

You're not alone if you say yes. Social psychologists are familiar with this effect, sometimes called "the man who" effect. I say: ninety percent of the people who try this drug receive benefit from it. You say "I know a man who tried it and nothing happened!" Somehow the face-to-face testimony of one person seems more convincing than any amount of statistics.

It might not sound very rational because those statistics represent lots of individuals just like your cousin. But putting a face to a problem seems to be enough to convince some of us.

Read more