That’s the headline you won’t read in this morning’s Independent, but it is the most accurate way to sum up their story claiming that a “Hung parliament beckons as Tories lead by four points”.
You might think this was based on a new poll, but presumably the Indy were to cheap to commission one, so have opted instead for the old political journalists’ trick of compiling a “poll of polls” instead.
This is an exercise that involves taking different polls from different times with different sample sizes which were then weighted in different ways, averaging them all out and then claiming that the average much be somehow statistically more significant than any one of them. Except that it isn’t.
It’s one thing for a polling company that uses the same methodology to aggregate different surveys over a period of time to create one big sample, which is then weighted in the same way, in order to get a more statistically valid result.
But it’s quite another to simply take an average of the topline figures of polls that had completely different samples – the results mean very little. And predicting the next election based on that at this stage is quite ridiculous. They might as well hire Mystic Meg as a pundit.
Unfortunately journalists too often get away with this as an easy way of filling up their pages so they can sneak off for a boozy lunch instead of doing any actual political reporting.
Next time they’ve forgotten to go and find any actual news, we suggest to the lazy Indy hacks that they go and play with their conkers instead of making up nonsense about polls.