Page 29 - ELITE PLUS MAGAZINE VOL2

gubernatorial election between the Democrat MR Sukhumbhand Paribatra and Pheu Thai party member Pongsapat Pongchaoren, pollsters mistakenly predicted that Pongsapat would win. In the too-close-to-call election, six of seven polling organizations forecast a Pheu Thai victory. Researchers know it is unwise to predict an outcome when results fall within the margin of error, but pollsters ignored this rule and the media reported their findings.Another common mistake is asking the wrong people. In the US presidential election in 1948, a poll made a poor sampling choice by getting responses mainly from Republican voters, thus falsely concluding that Republican candidate Thomas Dewey would beat Democratic opponent Harry Truman. The media reported the flawed poll results.Mr Hervio urged readers of polls to pay special attention to two aspects. “The ingredients of a reliable poll are a robust sample size and good, simple and unbiased questions.”A robust sample is a group of respondents that represents the population in question. An under- standing of Thai political opinion, for example, would require sampling from all the regions, genders and backgrounds that comprise the Thai people.Questions should not be leading or misleading, and must be unbiased and simple. Questions such as: “Is killing bad?” or “President Obama didn’t send troops to help Americans in Benghazi. Should he have?” are questions that lead and encourage predictable answers. The Gallup poll’s questions on levels of happiness may be overly complicated: Define your level of“Thailand has an insatiable consumption of poll data, especially in the news media.”happiness: 1) Are you very happy? 2) Are you fairly happy? 3) Are you not very happy?Mr Hervio thinks the Thai media should pay more attention to how a poll was conducted and inform the public of the details regarding data collection and questioning used in the poll.“Thailand has experienced political conflicts. We had a coup d’etat. We have heard and read a lot about public opinion, but there is no mention of sample size. There is no mention of when and how the poll was conducted – and this is not very good.”The Thai media has an infatuation with poll results. Political scientist John Grey Geer in his book Public Opinion and Polling Around the World wrote: “Thailand has an insatiableconsumption of poll data, especially in the news media.”Ms Daranee, who was formerly a research associate at Harvard Medical School, advised readers to be sceptical when reading about public opinion.“Do you know who commissioned the poll? A political party, multina- tional company, NGO? Anybody can commission a poll. You have to ask why. When I read about a poll, I always think about what’s behind it.”She suggested that the Thai media should follow the rule in the US of only reporting polls with a respected sampling method. The media should provide relevant information such as who sponsored the poll and how it was conducted.Elite+ 27