About The Sarah Palin Poll
The Sarah Palin Poll
From John Siceloff, Executive Producer, NOW on PBS
New and Improved - One User, One Vote.
I am writing in reference to the now famous "Palin poll" that appeared on the NOW on PBS homepage on September 5. It's no longer on the homepage; in fact, it's no longer a page that the online user can navigate to using menus, but hundreds of thousands of unique visitors are still accessing the poll and voting. How is that possible, and is that a good thing? More on that in a moment, after a side trip through the world of Internet "cookies."
PBS headquarters made the decision on September 22, to implement a cookie registration system on the Palin poll. That system is now part of the poll's inner code, and as of September 23, a user can only vote once per computer. PBS acted because the entire pbs.org site had been experiencing system overload due to massive accessing of the poll.
The poll asks the question, "Do you think that Sarah Palin is qualified to serve as Vice President of the United States?" The user answers clicks to answer "Yes" or "No." From September 5 to September 22, our software allowed online users to vote repeatedly. Sounds bad, right? But we at NOW had serious concerns about user privacy. "Cookies" are small text files placed on the user's system without that person's knowledge. Many computer users regard them as invasive.
Other possible fixes: we could have insisted on voters registering with their email addresses. But people could still vote several times using different email addresses. Or we could insist on a unique identifier - a user's social security number. This was clearly way over the line in violation of user privacy.
What are other media organizations doing? CNN.com does exactly what we had been doing—their polls allow multiple votes. The website of the newspaper USA Today also uses polls, called "USA TODAY Snapshot", but it employs the "cookie" approach to restrict multiple voting.
So, is the Palin poll now "scientific"? Absolutely not. It is still subject to large scale efforts on the left and the right to mobilize people to vote. The poll has become something of a Rorschach test, a tiny political marker in a tightly contested race. Over the past two weeks, the results of the poll see-sawed back and forth from a majority saying "No" to a majority saying, "Yes". At the moment the single-voter system was implemented, it was close to a tie: 50% say Sarah Palin is qualified to serve as Vice President, and 48% say no. Those results, in my view, are actually a measure of the mobilization and manipulation efforts by partisans on both sides. Now it will be all about mobilization, and less about manipulation. Blogs on the left and right are circulating viral emails with the exact address of the poll.
Some users have raised questions about our decision to collect opinion about Sarah Palin's qualifications in a poll. For a more complete discussion, take a moment to read the September 19 column of Michael Getler, the PBS Ombudsman.
The Palin poll is no longer in our home page rotation. We've moved on to other polls; each week you'll find one in the bottom right corner of our home page. The current poll asks, "Who do you trust more to fix the nation's economic mess—Barack Obama or John McCain?" It has already attracted a lot of interest.
And let's keep in mind the purpose of these weekly polls. They invite people who might not ordinarily come to our site to participate, and they encourage people to explore the deep journalism that is offered throughout NOW's nearly 10,000 web pages and 1,000 video streams. They don't provide the nuanced interpretation of our investigative journalism and deep content, but do provide a gateway to that content. An example: on Friday, September 19, we aired a one-hour special about women and politics on our broadcast. In the days since the air date, that online video has become the most viewed video on the PBS Election Site. For online users who are deeply interested in the issues raised during this election, the polls are only a starting point. There's a lot more to explore, to read, and to view.
Learn more about John Siceloff
UPDATE: As of October 15th, the Palin poll had received over 52 million votes. We've received inquiries about why the percentages in the poll results have not changed much in the last couple of weeks. There's a simple answer. There are so many votes now that even the hundreds of thousands of new votes each day don't have much of an effect on the overall totals.
Another question is about math. As of October 15th, the "Yes" answer had 49%, the "No" answer had 49% and the "Not Sure", 0%. The question is why these numbers don't add up to 100%. Here's the reason: the poll software rounds the results down to the nearest whole number. The "Not Sure" is less than one percent, so it rounds down to zero. The rounding-down is why why the numbers may not total 100%.
With all the questions--and answers--about the percentages and the math, it's easy to take a further step and regard the results as accurate and scientific. Not so! The Palin poll is a measure of voter interest and mobilization in the closing weeks of the campaign. It should not be taken as an accurate barometer of the views of the wider public about Sarah Palin's qualifications.