A regular absurdity of political opinion poll coverage in the UK is the way that media outlets often pretend that polls commissioned by other organisation do not exist.
If a polling company carries out, for example, two opinion polls in identical ways for two different newspapers, then the newspaper that publishes the second poll almost invariably ignores the first poll when publishing vote share changes, and instead bases them on the previous poll published in its own pages – even though that previous poll is older than the one published by its rival.
It’s as if at a general election someone reported the change in seats compared with not the last general election, but with the one before – just because they happened not to have covered the last one.
A round of applause then for Julian Glover and The Guardian, who left all this behind when reporting the latest ICM poll and instead make reference not just to the previous Guardian/ICM poll but also to the most recent ICM poll, which happened to appear in a different publication.
Now, to get a standing ovation someone would have to admit that a poll they’ve just published is a rogue poll which falls outside the normal margin of error for sampling (as you would expect around 1 in 20 to be – i.e. one about every other month in the UK as typically there are around 10 different polls with national voting intention figures published each month).
Follow @libdemvoice.org on Bluesky
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
Violet Bonham Carter and Mary Stocks remember the struggle for women's suffrage
Brighton Conference
Show us the money - should politics be funded by billionaires?
Jason Connor
Phil Wainewright
Christina Evans
Suzanne Fletcher
3 Comments
I thought you were going to say ’round of applause’ for this article about Vince: http://www.guardian.co.uk/politics/2008/feb/20/vincentcable.liberaldemocrats
The problem with your applause is that the ICM Guardian polls and other ICM polls are not consistent. The Guardian polls regularly give us a better rating than other ICM polls. This has happened since September and is causing comment elsewhere.
We all want to believe the polls that are best for us, but this is undermining my belief.
Halfway good point Mark. Different polls may have different questions, different interviewer protocols, different methods, different sampling models even if done by the same company. If they are to be compared then this must be taken into account in the analysis.