Scarcely a day goes by when the Hong Kong media are not highlighting the results of one poll or survey or another.

We learn that: women are becoming increasingly sexually liberated; Hong Kong is becoming a bachelor's paradise; up to 1.9 million people in Hong Kong have, or are likely to suffer from, eczema; and more than 60 percent of Hong Kong women regularly dye their hair to cheer themselves up.

The government conducts secret polls, as it did recently on how people feel about democracy.

In late June, Chief Executive Donald Tsang said Hong Kongers are more concerned with livelihood, the economy and almost anything else other than democracy, according to a poll taken by the government's Central Policy Unit.

ADVERTISEMENT

"My governance platform will be based on the poll results, same as my priorities," said Tsang, proclaiming the birth of government-by-survey.

He said 1,200 people were polled and the desire for universal suffrage ranked 13th on the list, far behind the top five concerns of unemployment, governance, air pollution, medical system reform and poverty.

But in that poll, neither the methodology, the questions nor the raw data were disclosed to the public, prompting one prominent local pollster to accuse Tsang of "survey abuse."

Writing in The Standard, Michael DeGolyer, of Hong Kong Baptist University, said polls in the territory are sometimes used as a substitute for democracy.

What is needed, he wrote, is "a governing system that uses a full array of market signals, not just `demand' as indicated by a telephone survey."

About the only topic that has not been a survey subject is polls themselves.

Do not rule that one out, said Robert Chung, director of Hong Kong University's widely respected Public Opinion Program which has been surveying Hong Kongers on political and social issues since 1991.

"We haven't done any survey on surveys, but I would love to see it done," he said. "Unless someone commissioned us to do it, any measurement of polling credibility could be seen as self-advertisement. This does not mean we will not do one in future for academic publication purposes. We just do not have any plan yet."

Politics plays a huge part in local polling, from self-serving organizations, the government and political parties to more objective researchers such as Chung and DeGolyer, whose Hong Kong Transition Project has tracked changes in attitude toward the SAR and Beijing governments since 1993.

Hong Kong's unique brew of language, dialects and religious beliefs makes truly accurate polls and surveys difficult, DeGolyer said.

Other Hong Kong opinion trackers use Cantonese exclusively but his group questions in Cantonese, Putonghua, English, Hakka and Chiu Chow.

Chung's POP sticks to Cantonese due to staffing and training limitations.

"It's simply a matter of resource constraints, and how much error you are prepared to tolerate," he said. "Using more languages requires more interviewer training and human resources, and may not be sustainable in the long run."

DeGolyer gets about 3 or 4 percent of his samples in English, Putonghua, Chiu Chow and Hakka. "Whenever you talk about surveying there are so many different aspects that need to be done right starting from the survey questions to the language or languages used in asking the questions and finding a truly representative sample," he said.

Birthdates can also throw Hong Kong opinion takers for an unwitting loop, DeGolyer noted.

To help ensure random results many pollsters who call multi-party households use the "next birthday" method of asking for the person who most recently had a birthday or whose birthday is soonest upcoming.

But there are many in Hong Kong who for cultural purposes count their birthday on the second day of the Lunar New Year, or are reluctant to admit it might fall on an "unlucky" date, DeGolyer said.

He sidesteps the birthday problem by using a "Kish table" which employs a random number chart to ensure that each eligible person in a household has an equal chance of selection.

It also helps with the convenience bias, which he described as "whoever answers the phone will say they are the person you want to talk with because they think it's inconvenient for you to talk with anyone else."

Roland Soong knows all about convenience factors in polls and surveys. He is the chief technical officer for KMR, the world's second-largest media research firm.

"If anyone were to call this number, my home number, it's a landline and it's listed in the book, there would be a problem," he said.

"Four people live there, three within the target universe of an adult. One is my mom who can't talk anymore. Another is a 79-year-old domestic helper who only speaks peasant dialect Shanghainese. So how's she going to do a interview about the chief executive? She has neither the verbal skills, knowledge nor interest.

"A Filipina maid is the third person. What are you going to ask her?

"I'm the fourth one. If you say you'll take `anyone' who is willing to talk and I'm willing to talk, I'm that person. But it's misleading for polls to say they project the total adult population of Hong Kong. They don't."

The other problem, Soong says, is that many pollsters only sample landlines. "I don't think they can sample mobile numbers and there are many people here with only mobile phones."

Chung, Soong and DeGolyer also noted that one of the quickest, most inexpensive and inaccurate methods for surveys is employed by political parties such as the Democratic Alliance for the Betterment and Progress of Hong Kong.

Numbers are dialed at random between certain hours and those who answer hear a recorded voice in Cantonese announcing the topic and asking the respondent to push a number to answer the questions.

Such was the method used between August 8 and August 18 this year when 1,506 people punched buttons for a DAB survey on minibus safety.

At a press conference on September 14, DAB councillors Raymond Yeung and Michael Luk solemnly announced the results to 13 local reporters: two from television, two from radio and nine from Chinese-language newspapers.

The results? Most people contacted have seen drivers speeding and/or running red lights despite precautions such as overhead speedometers and buzzers that signal speeding.

The DAB's recommendations? Based on the supposedly hard data of the survey, more riders should report bad drivers to the police, consideration should be given to installing equipment such as speed governors on minibuses and bad drivers should be prosecuted.

But not all of the percentages cited in the minibus poll totaled 100 percent and no margin of error was cited, details that did not seem to bother the reporters or their employers when the stories were broadcast and printed.

More careful pollsters such as Chung, DeGolyer and Soong use complete statistics and include a margin of error for their findings.

" I think 90 percent of what you see in the press are not examples of proper polling," Chung said.

"In the US they would just be dumped. The local media is doing a bad job as a gatekeeper. They encourage the publication of sub-standard polls, and they sensationalize poll findings."

DeGolyer said: "If you're trying to gather propaganda it's much easier to use recordings or give volunteers some phone books and have them make some calls. It's also where you can get a lot of results that aren't worth the paper they're printed on."