mental_floss magazine
SUBSCRIBE >
GIFT SUBSCRIPTIONS >
DIGITAL SUBSCRIPTIONS >
subscriber services >
Peter Donnelly is a statistician with a sense of humor. He starts his talk with a classic statistician joke: “How do you tell an introverted statistician from an extroverted statistician? The extrovert is the one who looks at the other person’s shoes.” But he’s not all fun and games. In a 2005 TED Talk, Donnelly explains a little about with it’s like to be a professional statistician, then launches into a fascinating explanation of how statistics are misunderstood by typical audiences. He gives examples for the audience to examine (a few multiple-choice questions), to prove his point — and I’ll admit, he got me. I didn’t get the examples right, though I was pretty confident in my answers.
This reminds me of when I was on a jury room in a personal injury case. While my experience wasn’t related to statistics, it was an issue of science which seemed like we should have been able to prove the right answer one way or another — but we failed. The jury had an hour-long argument about physics, trying to determine whether a driver’s upper body in a car would be pushed forward or backward when the car was hit from behind. (There was a key question regarding whether a specific injury could have been caused by the impact, or was a pre-existing condition.) We even built a model, but that failed to convince anybody of what the real-world behavior would be. Everyone on the jury insisted that his or her own mental model was correct (which tended to align with their gut feeling about the guilt or innocence of the defendant), and neither demonstrating the physics of the situation, nor thinking through it with a shared mental model made any difference. It was an interesting day, to say the least. (See also: whiplash.)
Anyway, Donnelly’s talk is a great example of how attorneys (or really anyone) can exploit general misunderstanding of statistics in order to make an invalid point — and most of us won’t notice that anything is wrong. There’s even a term for one specific statistical misuse, the prosecutor’s fallacy, and Donnelly explains how it was exploited in the Sally Clark case. Definitely worth a look!
TED Talks are awesome — everyone who loves Mental Floss should go to iTunes and subscribe. It’s basically a convention of very smart people and they invite smart people with big ideas or different ideas or just interesting ideas and stories to tell to talk to them.
posted by Southern Buddhist on 8-29-2008 at 3:40 pm
That was so interesting! I am totally subscribing to these talks!
I remember doing the coin toss experiment in my statistics class but it was not nearly as enlightening as this video.
posted by Eve R. on 8-29-2008 at 4:17 pm
In the economics book More Sex Is Safer Sex, the author argued that we should have professional juries. This talk gives further support for that case.
posted by Zach on 8-30-2008 at 12:53 am
Chris,
So…what was your jury’s decision and how did it arrive at it?
posted by Rich on 8-30-2008 at 7:43 am
I have to say - regardless of my personal opinion - I think that the sheer amount of statistics and information had an impact on the O.J. Simpson case. I think that the jurors were SO overwhelmed with the amount of data being thrown at them, that they didn’t have much choice but to return a not guilty verdict. The “reasonable doubt” dissipated with the massive number of “experts” they had to consider.
posted by Sandy on 8-30-2008 at 9:42 am
@Rich - well, long story short, I didn’t think the prosecution met their burden of proof — I thought the defense effectively refuted all the evidence presented, and I just didn’t see a case.
Having said that, the jury standard in Oregon allows for a conviction with a certain number of jurors disagreeing with the majority (I think it was 10 for, 2 against, or some similar number). I was one of those two who disagreed with the conviction, but it went through.
It was a little amusing — even though I disagreed with the majority of the jury members, I tried to help them talk through their issues and respect the rules as they worked to reach an agreement. It was amazing to me how often people directly ignored instructions from the judge (perhaps they just forgot?), and had to be reminded what was and wasn’t allowed. For example, many people thought we should go ahead and convict because the defendant seemed well-off, and presumably had insurance that would pay for the damages, which would give the plaintiff a better life, which they felt he deserved. (!!!) Oh well.
posted by Chris Higgins on 8-30-2008 at 11:54 pm