Thursday, August 28, 2008

Our Survey Said

I didn't really learn much myself, but I was quite impressed at this attempt at public education by the BBC, teaching people to deal with damned lies and statistics. It's part of a series, the most recent of which explains why most people earn less than average. (Sadly, I can't find a way to link to the series as a whole - it doesn't appear to have its own page).

In a world where numbers are thrown around to make sensational headlines, and their meanign rarely scrutinized, this is a welcome development. Ironically, it comes at the same time as the BBC report this regional happiness survey, glossing over the admission "researchers stress that the variations between different places in Britain are not statistically significant".

Worse, they an attempt to infer causation from correlation, and probably get it back to front. "Another important factor in determining happiness was the length of time that a person spent living in a place. Staying in an area for five years or longer was an important indicator of well being, said the scientists." The scientists point out only that length of time is an indicator of well-being, which to me makes it sound more like the dependent variable (i.e. people stay because they're happy), rather than - as the BBC report - a 'factor determining happiness'.

While I repeat my admiration of their attempt at public education, this not the only recent example of shoddy reporting - I pointed out a contradiction in this other piece before (and I have seen errors in their football reporting too, even if they didn't actually do this). I'm certainly less impressed by and more critical of BBC reporting than I used to be.

No comments:

Post a Comment