Should the Guardian check facts? (updated x3)

This morning, The Guardian’s website leads with a polling sensation. Apparently, Nick Clegg did so well is last night’s debate that the British election is now a two-horse race between the Lib Dems and the Tories, with Labour the UK’s new third party.

Such a sensational story – and so prominently placed – must be built on robust foundations, right? No. It turns out that:

  • The Guardian has picked up a tweet from Tim Montogomerie, editor of the blog, ConservativeHome.com, referencing an -as-yet-unpublished poll from ComRes for ITV.
  • Without even seeing the poll, it has decided to splash it all over the front page of its website, even while its own election blogger, Paul Sparrow, is warning readers that the sample is not nationally representative and the figures have not even yet been weighted.

This – apparently – is called journalism.

Update: Original headline pulled, replaced by the milder: “Lib Dem support surges after debate win.”

Update II: Andrew Sparrow writes:

I’ve been reading the comments about the way we reported the ComRes poll. Twitter is a wonderful source of information (if we didn’t use it, this blog would be slower and far less informative) and the figures were released by journalists who are normally reliable. But on this occasion the information was misleading. I should have waited until I had spoken to ComRes before going into overdrive. I’m sorry about that.

My beef is not really with Andrew’s live blog – which is clearly contingent and updated as new information is uncovered – but with the decision to use the information from a tweet as the basis for the main headline on the Guardian website front page. Would love to know which editor made that decision.

Update III: Here’s the Guardian’s description of the methodology for its own poll on the debate:

ICM Research interviewed a random sample of 505 people by telephone on 15th April 2010. It reinterviewed people who had previously been selected at random who told them they would be watching the debate and had agreed to be interviewed again. The sample has been weighted to the profile of all people selected at random who previously stated they would be watching the debate.

I have read this a dozen times and have no idea what it means.