Research company boss defends poll credibility

Listen to this article:

Research company boss defends poll credibility

CAZ Tebutt, managing director of Tebutt Research, has defended her company’s credibility in a statement issued from Australia yesterday.

“Tebbutt Research conducts all its opinion polls in accordance with the world code of practice published by the ICC (International Chamber of Commerce), WAPOR (World Association of Public Opinion Research) and ESOMAR (World Association of Market Research Professionals) and Tebbutt Research is the only Fiji member of WAPOR and ESOMAR,” she said.

Responding to the Attorney-General’s comment that the company’s research had not always been accurate, Ms Tebutt said the Tebutt Times pre-election polls were correct and called the result with a high degree of accuracy.

“The Tebbutt Times Poll immediately prior to the September election reported the results for all parties within the margin of error, except NFP which was 1.3 per cent points outside margin,” she said.

Ms Tebutt said with one in five people undecided or refusing in that latest poll, this level of accuracy was testament to the robustness and reliability of the Tebbutt Times Poll.

“The recent Tebbutt Times Poll on the Fiji flag was conducted in accordance with global best practice and Tebbutt Research stands by its methods and findings.

“This Tebbutt Times Poll canvassed the opinions of 1052 randomly selected adults from all walks of life across Fiji.

“Personal telephone interviewing was used, with both landline and mobile phone numbers being randomly selected, ensuring that both urban and rural people were canvassed.”

She said all responses were post-weighted to population statistics reported by the Fiji Bureau of Statistics so the results could be projected to the adult population of Fiji.

“Cross-tabulations were prepared to compare between different demographic groups of people (based on age, gender, location, urban/rural, and the ethnic group they identify with). Appropriate tests of statistical significance were applied.”

She said the maximum calculated margin of error in the total sample was 3 per cent at the 95 per cent confidence interval.