YOU DON'T TRUST POLLS? YOU'RE NOT ALONE

By Mary V. Coit

Polls are biased in the selection of the people they poll, in the politics of the group sponsoring the poll and even in the wording of the questions.

These are the views that surfaced in an independent survey conducted to find out the opinion of the public about polls. Voters say that political polls are not the scientific, objective measuring instrument they claim to be.

In answer to the question: why do you think different polls come up with different figures?

· 90% said it was because of bias
· 80% said it was the selection of people polled
· 50% said it was the agenda of the polling group
· 40% said it was the wording of the question

Other results from this recent telephone survey that show voters distrust polls:

· 8 out of 10 said a person would purposefully lie in answering a question on a poll.

The main reason for lying in a poll, voters said, was to project a positive image. Bending the truth to appear more important to others is a common reason people lie. This view is supported by a recent study at St.Louis University: researchers unexpectedly found that both men and women age 20-84 lied about their height, making themselves taller. Height is often equated with power.

Some people in this survey felt that respondents would lie in order to affect the outcome of the poll. Others said the urge to lie would depend on how seriously the respondent took the polls. A person might think that polls are stupid and lie as a way of messing with the pollster's mind.

Voters say they are interested but not influenced by polls.

· 7 out of 10 voters said that they pay attention to political opinion polls in newspapers and on TV

· All said they were not influenced by the polls

Most follow the polls by regularly reading the headlines. This attention to polls is a marked increase from the results of a 1985 Gallup Poll to find out about the public's awareness of public opinion polls: 41% said they regularly or occasionally followed polling results.

The voters in this survey refer to themselves as independent and freethinkers who would not be swayed by polling results. This independent attitude contradicts a 1992 NBC/Wall Street Journal report that 63% of those surveyed believed that voters were influenced by polls.

Answering the question: What do you think other people think about polls?

· 40% think that other people have a positive attitude towards polls; 50% think they have a negative attitude

Those who thought others had a negative attitude towards polls admitted to being cynical both about their attitude and the attitude of others. Those who had a positive attitude thought the reactions would range from being annoyed to thinking there was merit in polls. The general comment was that real people don't quote poll numbers, only those in the news media do.

The voters don't trust the way polls are conducted.

· All voters felt that the gender and race of the Interviewer can affect the response in questions about gender and race

· All voters believed that the question wording can affect the answer

Some felt that gender would be more likely to influence the answer than race; others felt the opposite. All agreed that these interviewer characteristics would have a strong affect on the answer.

Men will tend to give a positive answer to a question about equal pay for equal work if the interviewer is female but a negative one if the interviewer is male. Studies show that the gender of the interviewer affects the response not only on gender related question but others as well. This is also true of the race of the interviewer in racially sensitive questions. Although the respondents in this survey questioned how race could be identified on the phone, studies show that people can correctly identify the race of the interviewer three quarters of the time

Voters believe that the way a question is worded can lead to the answer the pollsters want. They know the wording of a question has an impact on the answer. Pollsters know this too, but don't tell the public that this is one of the major non-sampling errors that can contribute to the inaccuracy of the poll. The exception is The New York Times that recently has listed the wording of a question as a source of error that "may lead to somewhat different results."

Few believe the media polls

· 8 out of 10 voters don't trust the media polls such as ABC/Washington Post, CBS/New York Times, NBC News/ Wall Street Journal

The majority believed that the fact that the media conduct as well as interpret polls in the newspapers and on the networks could distort or give misleading results. Those who had faith in the media relied on the credibility of well-known journalists. Others argued that the media focused on the candidate they endorsed or what made a better story. The cynical conclusion was that media were not objective but rather biased in their stories.

Pollsters recently questioned the reliability of NBC News polling because the results of their opinion polls are based on a focus group of 36 "uncommitted voters" rather than the usual random sample. This is not scientific, the pollsters argue.

Although pollsters can claim to be reliable because of their scientific approach to public opinion, my survey shows that the public sees that there are too many basic problems with the method and bias of the pollsters for them to trust the results.

Pollster Patrick Caddell said polling "is an instrument of deception whereby the truth is obscured and the public will excluded and ignored." Voters in my survey would say "absolutely."

How this poll was conducted: This article is based on a telephone survey to find out what ordinary voters think of political public opinion polls. The survey was conducted from August 25th through September 2, 2000. The Respondents were not self-selected: they represent the demographic profile of Americans in terms of gender, race, education, income and age. The sample of 10 strangers is the size of most focus groups. The interviews took from twenty to fifty-five minutes and were recorded.

Mary V. Coit received her Ph.D. from Georgetown University in Linguistics and is a member of the American Association of Public Opinion Researchers. She lives in Washington, D.C.