One of the nice things about YouGov's research is that a lot of the time, they make the actual numbers from the research available.

This means that other people can lok at different ways of interpreting the data, and find new value in the figures - such as this piece of work from Election Calculus on how Tectonic Voter Migrations hurt the Conservatives;

Its an interesting piece of analysis, and also puts it into the political context (ie. the impact of Lib Dem voters switching on a likely coalition government.) Indeed, its an interesting election from this point of view – having grown up being told that anything other than a Conservative/Labour vote was "wasted", its looking like the key voters that will influence the next government are going to be exactly those that affect the smaller partners of a possible coalition.

As the article explains, the data comes from an anlysis of four YouGov surveys, including this one from the Sunday Times, where you can also look at the other questions being asked. Such as;

  • "If the Conservatives win the next election and form a government, do you think the overall level of taxation will rise, fall or stay about the same?"
  • Do you think Ed Miliband is too close to business, too hostile towards business, or does he get the balance about right?
  • Britain's current system of submarine launched nuclear weapons, known as Trident, is coming to the end of its useful life and will soon have to be scrapped or replaced. What do you think Britain should do when Trident reaches the end of its useful life?

Lots of interesting stuff that you could dig into and analyse, broken down by total responses, demographics (age, gender, social grade region), voting intentions and previous voting behaviours.

And then there are some other questions at the end, which include;

  • How often, if ever, do you whistle tunes?
  • Do your parents (or did your parents) often whistle tunes?
  • And do your grandparents (or did your grandparents) often whistle tunes?

It turns out that the Sunday Times had a short column about whistling on the 11th April, followed by a longer piece the day after, which included some numbers from the YouGov survey above.

Its pretty easy to see how this would have happened; someone had the idea of adding a question to the back of an existing survey, and it happened to be this one about voting intentions. But the upshot of this is that you can see the split of responses based on how the political responses were split. So for example, we learn that;

  • Lib Dem voters claim to whistle most often (28% of those who intend to vote in this election vs 25% average)1
  • UKIP voters are most likely to say that their parents 'often whistled tunes' (39% agree/44% disagree), while Lib Dem voters are the least likely (33% agree/50% disagree).
  • People from Scotland are most likely to 'whistle often' (38%), while Londoners are least likely (17%).
  • Young people are most likely to whistle 'often' (32% of 18-24 year olds, vs 18% of 60+ year olds), and are also the least likely to think that 'fewer people whistle tunes these days' (55% of 18-24 year olds agree, versus 81% of 60+ year olds.)
  • Of those who think that 'fewer people whistle tunes', 31% say that 'its because people have iPods and music players instead'. Lib Dem voters were the least likely to think this though - only 21% of them agreed with the statement.

There are other surveys on the YouGov site, if you dig around a bit. So, you could learn which parties supporters feel angriest about voting (UKIP), or are least likely to vote to express an opinion (Lib Dem.) Or more interestingly, how aligned Lib Dem and UKIP voters are with Conservative and Labour policies (both are closer to Conservatives on the economy, closer to Labour on education and the NHS), or which parties supporters are most likely (UKIP) and least likely (Lib Dem) to vote.

  1. Its worth noting that although the Lib Dem voters seem to stick out the most, they are also the smallest (and therefore most volatile) sample of the survey. Unfortunately the Greens and SNP aren't broken out in the results – I'm sure that those figures could have led to some interesting conclusions.