Online Polls: Basically Worthless

By: Steve Outing According to a recent poll found on the Time magazine Web site, 60% of respondents think that President Clinton should be impeached. A poll on CNN's Web site says that 54% of the respondents think Congress should begin an impeachment hearing. A poll of more than 200,000 people published on found that 73% think the president should leave office.

Hey, wait a minute! Those figures are diametrically opposed to results from polls we've been reading in the traditional press in recent days. A New York Times/CBS News poll finds that 53% of the public favors no action against the president and wishes the matter would be dropped. An NBC/Wall Street Journal poll finds that 58% think the matter should be dropped. What's the deal with these polls?

You may as well disregard the polls found on the Web sites I cited in the first paragraph above, for those are for all intents and purposes worthless. They're all online spot polls, where visitors to those Web sites are invited to submit their opinions. They are not scientific polls, by any stretch of the imagination. Respondents are self-selected, and most often there are minimal controls to prevent "ballot-stuffing" by individuals voting more than once by using different e-mail addresses. Too often, as the Clinton scandal polls demonstrate, their results bear no resemblance to true public opinion.

Harmful polls?

This brings up the issue of whether news sites should even be publishing spot online polls. Such surveys have long been a staple of news Web sites, touching on topics as weighty as "Should Bill Clinton be impeached?" down to "Should Elm Boulevard be turned into a one-way street?" The question that online news publishers should be asking themselves goes something like this: If our online spot polls have a high probability of being an inaccurate gauge of public opinion, aren't we doing a disservice to our readers by publishing them? What if our "Elm Boulevard" one-way street poll is a wrong account of public opinion, and the city council is influenced by it?

The answer for many online news editors will be that disclaimers are run with the results of online polls, warning readers that the results are not scientifically accurate because respondents in the polls are online users -- and those with online access remain a minority of the U.S. population. Let the reader beware. The trouble is, that theoretical Elm Boulevard poll might be the only expression of general public opinion that's available. We'd like to think that the city council is not so stupid as to let an unscientific local online poll influence its decision-making, but it's conceivable that a council member or a member of the public might bring up the poll during a meeting.

A better way to deal with online polls -- given their considerable limitations -- is to follow the example of the Washington Post, which on its Web site offers a feature called Poll Taker. Web users are invited to answer the same questions as respondents in Washington Post/ABC News polls, then their answers are compared to results from that scientific survey. (The Post/ABC polls are based on random telephone interviews of just over 1,000 adults.)

Many other news sites seem less concerned about the potential impact of the inaccuracy of their online polls. On CNN's "Quickvote" feature, a disclaimer is displayed prominently next to the results: "This poll is not scientific and reflects the opinions of only those Internet users who have chosen to participate. The results cannot be assumed to represent the opinions of Internet users in general, nor the public as a whole." Which begs the question, why bother running them at all?

MSNBC also runs online polls and what it calls "Live Votes." In a recent sidebar article written by MSNBC's Lori Smith, she explains to readers why the online poll results are so different from the NBC News/Wall Street Journal poll about the Clinton scandal. Smith does a good job of explaining why an online poll of 200,000-plus people is less accurate than a random telephone poll of only 2,000 people. Here's her explanation of why MSNBC nevertheless runs the results of those online polls:

"Why then, does MSNBC have Live Votes? For the same reason MSNBC invites users to send e-mail and participate in online chat sessions and the bulletin boards: to encourage you to express your views on the news with MSNBC writers, editors and other readers."

Election's coming

With U.S. elections fast approaching, we'll see a lot more online polls in the coming weeks. It's difficult to argue with the notion that many of them will feature results that have no resemblance to public opinion in the real world. I fear that such polls will do a disservice to the U.S. electorate. I don't believe that a mere disclaimer is enough, and the only real solution to this problem is to limit online polls on substantial issues to the Washington Post Poll Taker model. Take a stand and offer your readers only scientific polls results, while allowing your online users to take the same poll as those who got called in the random telephone polls.

Am I suggesting a blanket prohibition on online-only polls? No. The litmus test might be on the significance of the topic being polled. The Elm Boulevard example is one where real harm could be done if the poll is taken seriously. A poll asking, "Which do you like better: Monica Lewinsky's new hairdo or her old style?", is better suited as a quickie online survey.

New Internet technology that allows for instant polls is all well and good. The polls are a good ploy to get people to visit and revisit a news Web site. (Indeed, I have advocated their use in the past.) But the wildly fluctuating Clinton online polls point out how the use of online-only polls can serve to misinform the public. It's time to limit their use to appropriate instances.

(Finally, a tip of the hat goes to Roman Godzich of TotalNews, who noticed the discrepancies in several online polls and suggest this as a column topic.)

Meet the staff

In my last column, I cited some examples of how news Web sites handle contact information for their staffs. Another good example of a staff page is that of the Times Leader in Wilkes-Barre, Pennsylvania. Writes Web site general manager Jim Van Nostrand: "When we went online two and a half years ago, I distributed questionnaires to the staff with questions like 'What's the strangest thing you've ever done?' and 'What's the strangest thing you'd like to do if given the chance?' The bios and photos are voluntary; the telephone numbers and e-mail addresses are mandatory. In addition, our publisher mandated that all department heads participate."

(Finally, a tip of the hat goes to Roman Godzich of TotalNews, who noticed the discrepancies in several online polls and suggest this as a column topic.)

E-mail column delivery reminder

If you would like to get e-mail delivery of the Stop The Presses! column, there are now two options:

1) Text e-mail. I send out a text e-mail message containing a brief description of the current column, along with a URL link to the actual column on the E&P Web site. To receive these regular reminders, sign up here.

2) HTML e-mail If you have a mail reader that can handle HTML messages, have the entire column delivered to you whenever a new one is published. Sign up here.


Got a tip? Let me know about it

If you have a newsworthy item about the newspaper new media business, please send me a note.

To receive an e-mail reminder each time we publish a new Stop The Presses! column, sign up here.

Archive of columns
This column is written by Steve Outing exclusively for Editor & Publisher Interactive three days a week. News, tips, and other communications may be sent to Mr. Outing

The views expressed in the above column do not necessarily represent the views of the Editor & Publisher company


No comments on this item Please log in to comment by clicking here

Scroll the Latest Job Opportunities From The Media Job Board