Friday 8 May 2009

More or less

I enjoyed listening to More or Less over lunch just now.

Of course, while the programme says it "is devoted to the powerful, sometimes beautiful, often abused but ever ubiquitous world of numbers" I would argue that where it gets interesting it is really about information.

Today it included an item on the way that we respond, charitably, better to individual cases than big disasters. That we respond emotionally to the one-child-fallen-down-a-well but analytically, calculating, to thousands drowning in a flood (and give more for the one child). I'm not sure they considered the element of closeness - with the thousands being overseas - but the point is well made.

Maybe this is why it is so easy to say "I still believe that is was the right thing to do" after "despite, you know, that hundreds of thousands of people have died, and that is a tragedy"


Allan Jones said...

I agree that More or Less is very good. As you say, the issue under discussion is usually one of meaning rather than number. What do these survey results actually tell us? Etc.

The same broadcast raised the problem of weather forecasts stated as percentage probabilities. What does a 30% probability of rain mean? I've often wondered. It could mean many things. My sense is that most listeners interpret it as an index of the Met Office's confidence in its forecast.

David Chapman said...

Yes, if I remember correctly, they suggested that some listeners would interpret it to mean rain in 30% of places, but I wasn't convinced. I would have thought that those who understand '30%' at all would mostly interpret it as you suggest - which is surely equivalent to the 'correct' interpretation given on the programme: that it will rain on 30% of occasions on which that make that statement?