Don't mess with opinion polls

PUBLISHED : Wednesday, 01 February, 2012, 12:00am
UPDATED : Wednesday, 01 February, 2012, 12:00am


Opinion polls are playing an increasingly important role in public affairs. When conducted professionally, they can be very useful tools to gauge the community's views on topical issues and provide valuable references for formulating policies. But surveys can be tricky affairs. The questions asked, the timing, the sample size and methodologies used can all influence the results, either willfully or unintentionally. It is important that opinion polls are properly conducted to avoid misleading outcomes.

The premature release of the chief executive election survey findings by the Baptist University's School of Communication was therefore a matter of concern. It first disclosed popularity ratings that showed Henry Tang Ying-yen had narrowed the gap with Leung Chun-ying to 6.5 percentage points, and then amended the findings to a bigger gap of 8.9 in a low-key manner four days later. It was confirmed that the sample size was 1,000 people, but the first release only covered 800 responses. The university explained that it only wanted to release the findings ahead of Taiwan's election to secure better media coverage.

It is still unclear whether the full picture has emerged. At stake are not just the university's reputation and credibility. There are concerns that academic freedom may have been interfered with, following the revelation that Tang's aide had contacted the unit. This has prompted suggestions Tang's team might have been seeking to influence the timing of the release. A thorough investigation is needed to establish what happened. The influence of opinion polls can be powerful. It is not unusual to see surveys being used as a political weapon to advance certain causes. Political parties, non-governmental groups and businesses have long manipulated surveys to their benefit. The government, too, has a track record of manipulating opinion during public consultations to help with implementing controversial policies.

Unless the 1,200 voters on the Election Committee attach high importance to public views when casting their ballot, the candidates' popularity ratings are unlikely to have a direct impact on the outcome. But that does not mean the views of the community are irrelevant. The lack of a public mandate through the ballot box makes winning public support important for the future chief executive. That is why the slightest changes in survey results are of concern to Tang and Leung. It is also the only way for people to be engaged in the process.

When gauging the moods and views of the community, certain professional standards must to be observed. There should be no room for substandard surveys. Caution is also necessary when reading the figures. Misleading findings only do the public a disservice.