As long as the method used and margin of error were spelled out, opinion polls were worthwhile, academics said at RTHK's City Forum yesterday. And they dismissed claims that being a political commentator conflicted with the task of survey research. Li Pang-kwong of Lingnan University said he did not see any conflict in a pollster commenting on his findings. 'It would not be a problem as long as there are objective principles and he is willing to face criticism openly,' said Dr Li, who heads Lingnan University's opinion poll and research programme. But he added that he thought research simply asking whether the public supported the Chief Executive had little academic value. The comments follow concerns triggered by the Robert Chung Ting-yiu controversy, in which the Hong Kong University pollster was accused of having a conflict of roles in being both a researcher and a political commentator. Dr Li said he believed opinion polls on politically sensitive issues would not be scaled down as a result of the controversy. 'A university has its own role to play,' he said. 'It may not necessarily support the Government, or oppose it. Whether a pollster should ask a question should be a matter for his judgment, that is, whether it is in the public interest to ask such a question or whether society needs to know about it.' So Moon-tong of the Polytechnic University said conflicts were bound to exist because different people had different roles. 'It all depends on one's academic ethics and quality, and whether one will stand by a set of guidelines impartially,' said Mr So, project associate of the computer-assisted survey team of the university's centre for social policy studies. The managing director of Hong Kong Polling and Business Research, Hung Ching-tin, said opinion polls were worthwhile for reference if the margin of error and methods were made clear. Professor Li urged the media to publish background information such as sampling methods and percentage of errors when reporting survey results. The researchers said the popular method of random telephone survey had limitations, because not everyone had access to telephones, while households with more than one telephone number had a bigger chance of being selected. The increasingly popular tool of gauging public sentiments through self-completed surveys on the Internet was also said to be unacceptable.
