Are surveys accurate?

We are all approached during the year by either telephone or face-to-face surveys being conducted by organisations such as A C Neilsen McNair on behalf of nameless groups or companies wanting to find out how we feel about things.

These surveys are treated by the public as an "accurate" reflection of how we feel about an issue, political party or product.

In the scanned image linked here you will see a letter from the Caboolture branch of Pauline Hanson's One Nation. The letter, signed by the branch President and Secretary, states: "One of the questions asked (by AC Nielsen McNair in the Burpengary area) is: 'If an election was held tomorrow who would you vote for?' They then show a list of political parties. Pauline Hanson's One Nation Party is not on the list."

A few phone calls to A C Nielsen McNair resulted in the following information being obtained.

The Queensland office confirmed that the survey was called a "National Omnibus Survey" which was headed up by Mark Sumich at the Sydney Head Office. They told me that the survey was based on 1,000 people met face-to-face and is based on a set number of people being interviewed in each state.

I then contacted Mark Sumich at the A C Nielsen McNair office who told me that the single survey was done for a number of clients and that his only client requiring political information was none other than Fairfax's Sydney Morning Herald. At this stage he made it quite clear that all questions in telephone and other surveys were used for assessment.

When I read him the letter we had received he said that "(The political question) was to get people interested in the survey and that the information is not used at all for the surveys."

I asked him about the question regarding voting intentions and he said that the list of political parties shown to people on a face to face basis had not been changed since January this year - before the launch of Pauline Hanson's One Nation.

He said that they should have One Nation on the list but they don't as they see it as a minor party and that it was catered for anyway by the list of political parties shown to people which he confirmed read as follows: "Australian Labor Party, Liberal Party, National Party, Democrats, Greens, Independents and Others". His comment was that people could show their preference by nominating One Nation under "Others" as their voting preference.

I asked him how The Sydney Morning Herald got its voting intentions survey results from A C Nielsen McNair and he told me that that was through a separate phone poll undertaken weekly.

While I do not question the accuracy of Sumich's statements it has got to be of concern that One Nation,  which has voting preference results all over the place depending on the survey being taken and by whom, is left off what is canvassed to face-to-face respondents as a legitimate survey for unknown organisations.

When you learn that the only client using the National Omnibus Survey for voting intentions is Fairfax's Sydney Morning Herald who report very negatively on Pauline Hanson and One Nation suspicions can only be aroused.

The discrepency in polling on voting intentions as they relate to Pauline Hanson's One Nation with a recent Queensland poll in the Courier Mail showing 18% primary support state wide while The Australian's Newspoll claiming just 2% at the same time! Sumich told me that a recent poll undertaken for The Sydney Morning Herald had shown a 7% support for Pauline Hanson's One Nation.

Could this, under the circumstances, be a remarkable level of support and confidence shown by ordinary Australians in Pauline Hanson's fledgling party?

Piers Ackerman on the credibility of The Sydney Morning Herald's surveys.
Return to the Australian National News of the Day