Interpreting the political polls in the final week
We tend to suffer from poll fever in this part of the election cycle with individual news organisations falling over themselves to bring us breaking news of the latest snapshot to reveal what voters are thinking. All stand by the accuracy of their own poll and tend to regard it as the oracle, neatly ignoring polls published by their competitors. This is not very helpful to the interested observer who is trying to figure out what all this means.
Political polls present a number of problems for those trying to interpret them. They all use slightly different methodology, there is a necessary delay between the questionnaire going out into the field and the results being announced, and they are just a snapshot in time.
Political scientists prefer to talk about trends over time rather than the significance of a single poll result – which introduces the concept of a ‘poll of polls’. This is where individual poll results are averaged out and tracked over time and is probably the most useful way to attempt to discern meaning from political polls.
Curiablog is currently doing a sterling job demystifying the polls. It’s running an up to date public poll average which weights the poll results according to factors such as sample size and the date it was taken.
This is a good example of the blogging fraternity filling a hole left by the competitive sensitivities of the mainstream media and doing all of us a favour by making sense of the numbers.
Tags: election, news, political poll