Interpret GM polls with caution: NZ academic

Polls claiming to illustrate community views on genetic modification should be interpreted with caution because they did not always take into account a person's knowledge of the topic, a New Zealand academic has warned.

Massey University survey research expert Assoc Prof Janet Hoek found that respondents often misunderstood the terminology of questions and did not necessarily grasp the concepts put forward.

She said that while this problem could be applied to various topics, it has recently been visible in the many GM surveys being conducted in the lead-up to New Zealand's general election.

The nation's Greens Party commissioned a poll of almost 500 New Zealanders that found 64 per cent of people said they thought GMOs should remain contained in laboratories. It also found that 24 per cent disagreed with the question and 11 per cent were undecided. On the question of GM crop farming should be permitted in the nation, 46 per cent said no, 35 per cent said yes and 17.5 per cent they didn't know.

By contrast, a National Business Review-HP Invent poll showed 76 per cent of Green supporters backed the outcome of the country's Royal Commission, which stated New Zealand should proceed with caution on GM, and a Herald DigiPoll concluded that 66.7 per cent of people thought GM organisms should be commercially released, after a detailed inquiry and under strict conditions.

Hoek said the latter poll failed to explore what respondents understood "commercial release" to mean.

"For some people this may mean field trials, others may believe commercial release means supermarkets will soon be stocking products containing genetically modified ingredients," she said. "Clearly, people's response to this issue may vary considerably depending on how they interpret key terms. Terms such as 'a detailed inquiry' and 'strict conditions' are ambiguous and respondents can interpret them in different ways."

Similarly, she said polls reporting that respondents favoured the Royal Commission's recommendations on GM needed to explore people's knowledge of these recommendations.

"Without some discussion of respondents' knowledge of key terms, it is not clear whether their answers are based on correct interpretations of the concepts being explored," Hoek said.

She said previous research by her department into the referendum question put to voters in the 1999 General Election found that some respondents misinterpreted terms such as 'minimum sentences'.

Hoek said the finding illustrated that even apparently straightforward terms could be misunderstood. "The potential for misinterpretation is arguably even higher with complex issues such as GM," she said. "Poll reports that do not consider these issues, need to be interpreted very cautiously."