Richard Willis's Blog

First for local news and first for comment


Sky News are leading on a “poll” which they gleefully report has the Lib Dems down at 8%, with Labour on 24% and the Conservatives on 43%. The reporting majors on the Lib Dems being under 10% and I looked to see the detail as to which credible organisation could have conducted the poll, what weighting was applied and the size of the sample surveyed.

The good news is that it is a respectable sample of over 1,500. However, that is where the credibility ends. This is a “poll” of the Sky News “Panel”. There was no weighting applied, leading to the ludicrous situation that 17% recalled as having voted Labour in the last election (when in fact they achieved 29%). Thus 24% is arguably an improved position for Labour.

The figures also include 13.5% of “dont knows”, “didnt vote” and “not eligible to vote”, which are normally stripped out of the final figures by any reputable pollster. This would boost all the major party shares if the figures were rebased in the normal way. I am not going to attempt to do this but I am sure some reader  can do so!

My best advice on this poll is ignore it. It is what we call a “voodoo poll” which appears to have as much to do with gaining a sensational headline as informing the political debate.

UPDATE: The well regarded UK Polling Report now has a comment on this Sky News “Poll”. Whilst he says that it is not strictly a “voodoo” poll, as I called it, he is scathing about the methodology and supports my advice to ignore it!


August 17, 2010 - Posted by | Polls


  1. You would have thought that Sky News would know better than to promote this crappy poll so heavily, or would at least have got the data sorted into an more acceptable format.

    Comment by Doodlebug | August 17, 2010 | Reply

  2. The 3 main parties add up to 75% . I’m surprised that you even bothered with it.

    Comment by Howard Thomas | August 17, 2010 | Reply

  3. Sky have never let intelligence or factuality get in the way of a headline, but thanks to them for providing a paradigmatic example of bad “analysis” – useful for anyone doing statistics at GCSE.

    Comment by Jonny | August 18, 2010 | Reply

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: