The Art of Asking Questions

An interesting pair of polls appeared over the weekend that intended to reflect public opinion on the Government’s policy on Libya. The first, by YouGov, suggested that 45% of people thought that “Britain, the US and France were right to take military action against Libya”, whilst 36% though they were wrong. The second survey, by ComRes, found that 35% of people thought that “Britain was right  to take military action against Colonel Gaddafi’s forces in Libya”, but 43% thought they were wrong.

Both polls were conducted at the same time and had around 2000 respondents which should mean that they accurately reflect public opinion. So how can the two be so different?

The difference comes from the way the question was asked and also the context of the question. The YouGov question asked not about Britain alone, but highlighted the fact that the operations in Libya were part of an international coalition. The ComRes question talked only about Britain, and also followed on from a series of earlier questions discussing the British economy and the latest round of spending cuts. Within the context of a troubled economic climate and the need to save on the cost of many public services, fewer people saw the value in committing scarce resources overseas.

Interestingly, it’s fair to assume that in mentioning Gaddafi, the level of support for action in the ComRes survey might have risen (as it did with similar surveys focussing on  the British involvement in Iraq that explicitly mentioned Saddam Hussein) and had ComRes talked only about Libya, the level of support may have come back lower still.

Neither survey is wrong here. Each asked a question and got an answer which does, in all likelihood, reflect public opinion on that question. The problems may come when the responses are subsequently interpreted and the specific questions asked begin to get lost. It’s a nice illustration though of how the phrasing and structure of a survey question can affect the results and perhaps encourages us all to look beneath the headlines to find out what exactly was asked when reading about survey results.

Just out of interest, how do you think the articles below presented the results of these two surveys?

Related Articles