Statistics and Juries

Today’s post is an incredibly wonky one–many of you will bail right off the bat, and of those of you who decide to check out this video, several of you won’t make it past the 2 minute mark. So what is the fascinating offering I’m posting? It’s actually a session from TED that I find to be really interesting: how statistics fool juries. If you don’t feel like you can commit to the full video, skip ahead to the 14 minute mark, where the reason for my posting the video becomes clear.

I think what makes this piece so relevant is actually two-fold: first, the presenter (Peter Donnelly) discusses how misrepresentation or misinterpretation of statistics can have brutal consequences in criminal trials (in this case, a murder trial in Britain). And he looks at the issue not just from the perspective of experts erroneously presenting the stats, but also the lawyers who never question what the experts are presenting. The second thing that makes this piece so relevant, though, is an even more critical concept: understanding professional limits. What happens when experts over-reach and draw conclusions about stats that are inaccurate and/or irresponsible?

I’ll be curious to know what people think. If nothing else, enjoy Dr. Donnelly’s whole Hugh Grant-esque accent, which I found to be quite pleasing at 1am when I was listening to this.