Numbers Game: Why Polls Have No Soul


November 07, 1991|By ALICE STEINBACH

I have to admit I felt very angry when I heard it on the radio a few months back: A survey that revealed that 56 percent of its male respondents said they'd trade their mates for a million bucks.

But before you decide to react to this bit of news as I did -- with anger -- a word of explanation is in order: It was not the message that angered me, it was the messenger. And the messenger's sources.

Or to put it more precisely: I've had it up to here with polls and surveys -- both the "unscientific" ones such as the call-in television surveys and the "scientific" ones that give us numbers that, when probed for meaning, often mean nothing.

For example, the "survey" that produced those "statistics" about males choosing money over mates turned out to have been conducted by 7-Eleven convenience stores. The methodology? Customers purchasing beverages at 7-Eleven were asked to respond to the question "Would you trade your mate for a million bucks?" by taking a cup marked "Yes" or "No."

The "poll results" then were passed along to the media, which passed them along to the public. Harmless fun? Maybe. But since there is evidence that a number of people believe all polls -- no matter who conducted them -- even a pseudo-survey such as 7-Eleven's can contaminate the cultural environment.

Worse yet may be the so-called scientific polls. For those among you who accept the notion that polls conducted by professional pollsters and university researchers offer an accurate reflection of public opinion, may I suggest you take a look at Susan Faludi's new book, "Backlash: The Undeclared War Against American Women."

Among other things, Faludi, a reporter for the Wall Street Journal, rigorously researches and takes apart several of the "statistical" developments that affected the professional and personal lives of women in the '80s.

Remember, for example, the 1986 Newsweek cover story that declared single women in their 30s "are more likely to be killed by a terrorist" than to marry? That now-discredited study did not come from 7-Eleven cups; it came out of a Harvard-Yale study. It was challenged almost immediately by a female demographer in the U.S. Census Bureau -- whose own study produced dramatically different results.

But true or not, the original study had a profound effect on single women over 30. Faludi reports that, as a result of this study and the news media's widespread coverage of it, many single women in therapy became " 'obsessed' with marriage, ready to marry men they didn't even love, just to beat the 'odds.' "

But gender issues and cultural attitudes are not the only areas in society that are constantly being measured by polls and statistics. Public opinion polls trying to gauge the political temperature of voters are taken on an almost daily basis and sometimes seem to threaten the entire political process.

So many dangers lie in the minefields of all this polling, so many questions need to be asked: Do polls measure public opinion or do they create public opinion? Do we rely too much on polls to direct our agenda? Do polls that show overwhelming numbers on one side or another become self-fulfilling prophecies?

And worst of all, perhaps, is the effect that polling and numbers may have on our ability to form our own opinions. Let's face it. Although we like to think otherwise, ours is not a great age for individualism. More and more we abandon the hard work of independent thinking to fall in line behind the presumed voice of the majority. Or the minority.

What we want to know before we form an opinion increasingly seems to be: What are the numbers?

Even pollsters agree that numbers are only numbers. Slippery things that are the result of, among other things, the way questions are asked, the order in which they're asked and the day on which they're asked. But people believe in numbers.

"Numbers seem to be truer than words," researcher Daniel Yankelovich once told Time magazine. "There is a patina of validity to a number that may not really have it. The result is a promiscuity of polling, a mindlessness. Nobody cares what goes in, so long as a number comes out."

Pollsters, of course, know that numbers can be corrupted. And polls based on numbers can be used in manipulative ways.

I wonder, for instance, how the Senate would have voted on the confirmation of Clarence Thomas to the Supreme Court if no polls had been taken around the Hill-Thomas hearings. Did the opinion polls showing wide margins of support for Thomas convince some senators that they could vote for Thomas without worrying about alienating women voters and jeopardizing their political careers?

We'll never know. Although I guess we could take a poll to find out.

Baltimore Sun Articles
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.