Wednesday 3 March 2010

a quick look at the wording of YouGov's latest opinion poll

As someone who takes a keen amateur interest in market research, like many thousands of people I take part now and again in online panel surveys through YouGov (the going rate is £50 every few probably works out at about 5p an hour or something!) Today I was invited to take part in an opinion poll, and as the polls are getting hotter and hotter as we approach the election, clearly the researchers are trying new methods of trying to ascertain voting intentions. The survey today asked a number of questions in interesting ways, so I thought I'd write a quick note on it while it was still fresh. Unfortunately I only started to make any notes o the questions when I was halfway through the survey, so my memory of the early part is pretty dim!

After beginning with the usual "what party will you vote for/who did you vote for in 2005" type questions, we moved to some sharper questioning. Much of it contained "weasel words" and leading questions, but I imagine quite deliberately to test whether a certain party political message was getting through more effectively than another.

One portion of the survey focussed on the economy. One question very pointedly referred to the "standard of living" potential PM/Chancellor combinations would provide - in 2010 I imagine more people in the country could name the Chancellor and Shadow Chancellor than any time in the last few decades. This was followed by "Which leader is best equipped to lead Britain out of difficult economic times?" I wonder if there will be much difference between the two questions; by taking the difference between the two indices we might get a picture of which Chancellor is more highly regarded (the survey didn't ask this explicitly, but perhaps there would be too many DKs. I find it hard to judge what Gordon Brown's economic reputation is these days - in the past he was regarded as an economic god, more recently an economic disaster - I suspect many people will see him somewhere in between.

One or two of the questions were quite subtely leading thanks to the choice of words used. "Who is the most honest about the economic problems facing Britain?" is always likely to induce bias against an incumbent's a damn sight easier for a challenger to be "honest" about problems that are of no concern to them at the moment! On the other hand, I felt that "Who best understands the problems ordinary people face in difficult economic times" contained a weasel word that would likely favour Labour. It'll be interesting to see the results. Most fascinating of all was "Who makes you most optimistic about Britain's future?" - perhaps the question every voter should be asking themselves in the booth come polling day.

A Likert scale with several statements heavily loaded with bias was then presented. "It is time for a change in government" obviously alluded to the Tories' election slogan, countered with "The economy is too fragile right now to have a change of government in the enxt few months". There was a cheeky comparison of David Cameron to Thatcher, juxtaposed with "Britain will be fairer under a Labour government". Elsewhere in the survey the word "lightweight" was used - a term which has been used on numerous occasions about Davod Cameron and which I feel he still struggles to shake off.

Presumably pollsters deliberately stack up their questions with bias but that doesn't mean some of the questions can be error free. A question asking who would be best at "tightening controls on immigration" makes the assumption that makes immigration controls tighter is in fact best for the country; for many a bleeding heart liberal (*waves*) this is a nonsense. I also wonder where the next question on "select your top three issues from this list" picked their issues from; tired old chestnuts like Iraq/Afghanistan and Europe, barely covered in the media in recent months, were beside all the major topics (tax, economy, education, pensions) but not some of the major current issues (corrution and electoral reform for example). Why didn't they take inpiration from the albeit massively unscientific Power 2010 poll?

Finally there was a comparison of Brown vs Cameron on personality issues. We were asked to select traits which applied to Brown from a list - these included "strong leader" and "can work effectively with colleagues", which were fixed like an anchor as they also applied to Cameron, and traits specifically for Brown ("bully") and Cameron ("just a salesman").

It was a longish survey to run over lots of similar questions and I rather suspect that were I not interested by either surveys or politics then my attention would have wandered as there was lots of similar wording. It's lucky I'm into both!

As someone who has never worked as a research professional, I feel rather shy about making a critique of a professional survey. One thing's for sure - I'll be fascinated to see the results and compare them to my predictions.

***Update 4 March*** It turns out that this poll was for Channel 4 in marginal seats (I live in Hampstead & Kilburn) and those nice people at C4 have uploaded a full spreadsheet breakdown of the results. Were my predictions right? I'll have a look tonight!

No comments:

Post a Comment