Can public trust the pollsters?

How much faith should the public have in polling evidence?

Figures are often cited by the media to suggest support for opposition to politicians, parties or policies, despite the snapshot nature of the data.

But when a poll is commissioned by a particular group or individual with a vested interest in the results should the public simply dismiss the survey findings?

This was the suggestion of Business Secretary Vince Cable over the weekend.  Quizzed on Sunday's Andrew Marr Show about a poll suggesting significant opposition to the planned VAT rise by supporters of Mr Cable's Liberal Democrat party, the Secretary of State quickly dismissed the findings.

"The poll you're quoting was cooked up by some researchers for Ed Miliband — it's not an opinion poll," he said.

However given that the poll was conducted by YouGov on behalf of Mr Miliband it would appear that it should not be necessarily be dismissed so lightly.

While groups or individuals can commission polling, if the pollster is a member of the British Polling Council, it will have to disclose the full data from its research, thereby making public any questions that might skew answers.

Likewise pollsters that are members of Market Research Society (MRS) have to subscribe to that organisation's code of conduct, which includes guidance on keeping questions objective rather than leading respondents to particular conclusions.

So while a group may commission a poll with particular questions in mind, there will often be some give and take between the group and their pollster of choice.

As Anthony Wells, Associate Director of YouGov explained: "In the vast majority of cases clients are happy to accept the professional advice of the researcher. In most of the rest of the cases, a happy compromise is possible — there are very few cases where clients actually want to skew answers, mostly they are just not versed in writing good, fairly worded questions

"The ultimate sanction is that the pollster refuses to conduct it — polls from pressure groups, etc, bring in very little money compared to big corporate clients so pollsters wouldn't risk their reputation for polls that bring in relatively small amounts of revenue."

While not commenting on claims on particular polls, Mr Wells suggested that so long as the poll data is published there is little grounds to dismiss a poll because of the people who commissioned it.

"This is the very purpose of the British Polling Council disclosure rules — if pressure group A claims they have conducted a poll showing X, you can get the actual tables off the pollster and find out that it is a rather idiosyncratic interpretation of the poll, and that fair minded people might conclude it shows Y.

"People who dismiss polls without actually looking at the tables and saying what they think is wrong with the research, just because X commissioned it are, to put it charitably, making themselves look a bit silly."