The sample of one

Many years ago the BBC radio series "Radioactive" - based on a very shallow local radio station - had a feature on a by-election. I am quoting from memory here, but the reporter said something like this:

"We asked Mrs Joan Smithers, of Mercer Road, how she was going to vote, and she said 'Conservative'. If that proves representative of the constituency as a whole, we predict the Conservatives will win with approximately 100% of the vote. If that were reproduced at a general election it would produce a Parliament with 650 Conservatives and no seats for any other party. If we project that into the future it would last for all time".

Of course, those were the days when the BBC used to mock absurdly unrepresentative polls, instead of reporting them as 'news'. I think it was the day before yesterday that the BBC featured an interview with an Iraqi 'housewife' - the word is in quotes because her house had just been destroyed - and, in all seriousness, asked her if her life was better or worse than before the liberation in 2003. She said it was worse. What does that tell us about the general situation in Iraq? Well, ummm, nothing.

There IS one generalisation you can perhaps legitimately infer from the BBC's groundbreaking interview. People who have had their houses destroyed tend to be less than pleased about it. I realise this is also drawn from a sample of one, but I am confident that a scientific poll of a representative sample in this group would confirm this. Congratulations to the BBC for uncovering this surprising truth.

View print friendly version

All information © copyright Quentin Langley 2019
RSS 1.0 Feed