So writes today's Guardian:
Britain's 5.5 million Twitter users are younger than average members of the public, slightly more likely to vote Labour, and distinctly more liberal, according to a survey published today.
They are also more likely to live in London, less likely to live in the north of England, and marginally more likely to belong to a lower social class.
The YouGov poll may help to explain why libertarian campaigns appear to flourish on Twitter.
You can read the full story here. Watch out when you do as it becomes clear that the evidence doesn't quite fully stack up given the margin of error on the Twitter user results is +/- 7 percentage points and that of non-Twitterers is +/- 2%. That means many of the comparisons in the piece stack up, but some don't - those about London, Conservative voting and social class all fall well within the margins of error and so the poll doesn't actually tell us whether Twitter users are more or less like to be Londoners, Conservatives or in the lower social class classifications.
- The coalition agreement: civil liberties
- Wrong ballot papers supplied on polling day in Brent
- More is not always better: how many Twitter followers do you want?
- When is a late surge not a late surge? (aka how Reuters gets the news wrong)
- Political myths questioned: Twitter is no good for local politics
As you well know, Mark, margins of error are extremely complex mathematical issues and you really can't expect journalists to understand all that scientific mumbo-jumbo. Good grief - next you'll be telling us they should be able to understand graphs or averages.