« Kristol Is Boring | Main | Guest Blogger Responds With Conspiracy Theory »
January 14, 2008
How Many Polls Does It Take to Change a Light Bulb?
That old, ethnic, politically incorrect joke setup was meant to demean the intelligence of people of Polish decent, but it works just as well spelled “polls,” because some uninformed, stereotyping people want to scapegoat the polls over the bad call in the New Hampshire primaries.
It’s just another case of killing the messenger. Even the old warhorse, Tom Brokaw, whupped up on the media and the pollsters for calling the outcome wrong. Well, it’s time for the contrarian Media Curmudgeon to stick up for the pollsters.
First, the media, especially the broadcast media, don’t bother to give the sample size and margin of error of poll results because viewers wouldn’t stand for it. Can you image how fast people would reach for their remotes if a TV anchor said, “The latest results from the Mark Penn Company polls of voters in the New Hampshire primary are in. But before we tell you the results, you should know that the sample consisted of 1,100 telephone interviews …” Thirty percent of the audience – CLICK. “…weighted for certain demographics, based on a random sample of non-cell phones, and with a margin of error of plus or minus 4 percentage points…” Fifty percent of the audience – CLICK. “…the poll taken between 10:00 a.m. and 8:00 p.m…” The remaining 20 percent – CLICK. “…yesterday shows that 38 percent of those responding indicated they would vote for Senator Obama tomorrow, if they voted, and 30 percent would vote for Senator Clinton, if they voted,” as the anchor talks to only the unmanned camera – the cameraperson having left with the sports anchor for a Bud Light.
Second, the pollsters merely reflect what people tell them during the time a poll is taken. They are the messenger for what people tell them. They can’t interpret what people really mean, if they are lying, if they are giving what they perceive to be a socially acceptable answer, or if they are saying something just to please or throw off someone who is listening to them in the room in which they are talking. In other words, the pollsters don’t predict an outcome, they merely report on what people told them at a given time with a margin for error, which no one pays any attention to.
The media is lazy and just spouts a headline, without any explanation or caveats. The media outlets are also gutless because they know that if the predictions are wrong, as they were in New Hampshire, that they can blame the pollsters and not take any responsibility for being wrong.
The headlines Tuesday night and Wednesday morning after the New Hampshire primary should have been, “Voters Change Their Minds!” or “Voters Lie to Pollsters!” or “It’s a Statistical Tie!” or “Obama and Clinton Tie For Delegates!”
Get real. Polls are not going away and the media are not going to stop reporting on the polls – people are dying to know what’s going to happen; it’s human nature to speculate on an outcome, even to bet on it. For example, check out Intrade Prediction Markets and see what the latest odds are for Clinton and Obama. Also, the polls are not going away even if no one responds, as Arianna Huffington naively recommends.
People aren’t going to stop watching weather reports because all the weather bureaus and TV weathermen said there was going to be a big snow storm today (Monday) in New York and it just sprinkled a little. The weather forecast was dead wrong, but we’ll keep listening to the forecasts because we want to know what the weather’s going to be, even when we know the forecasts are wrong occasionally.
If I were a gambling person, I’d set up an off-shore betting site – it’s against the law to have a betting website in the U.S. – I’d do one for betting on how accurate the political polls were going to be. I make a lot of money because people would bet against the polls after the New Hampshire error, but the polls would be right much more often then they were wrong.
So, how many polls does it take to change a light bulb? Just one, dummy.
Posted by Charles Warner at January 14, 2008 11:53 PM
Comments
Media Curmudgeon ![[TypeKey Profile Page]](http://www.mediacurmudgeon.com/nav-commenters.gif)
at January 15, 2008 06:34 PM writes:
Bruce Braun writes:
"Perhaps the roots for all of this polling hyperbole got its genesis in the arenas of sports and entertainment. It all comes down to box- office.
As a nation, we have always been obsessed with sports and the stats surrounding it. It starts with win/loose and then degenerates into endless numbers of team and individual performance. How many people do each of us know who are walking databases of sports stats?
Going back, 20-30 years ago, did anyone outside of Hollywood know or care about what movie was number one at the box-office in any one week? Today, reporting that information is standard fare for 99% of broadcast and print media reporting. The media treats box-office, any box-office number, as an indication of popularity, or in another more informal way: popularity polling.
What we are experiencing today is nothing more than political box- office every week.
No one, as they say, wants to be on a loosing team or spend money to see a lousy movie. To counter that thinking, carpet bombing of candidate TV ads are all produced to make us feel that voting for candidate X makes us a smart, insightful, and compassionate
individual. Now, add the top-heavy coverage of any candidate who
surges in a given week, and the news media effectivly uses the polling
as just another tool to generate viewership. The broadcast networks
and their partners in hyperbole, the print media achieve this by continuous network/newspaper "exclusive polls". Margins of error, the specific questions asked, the order of the questions, are as you point out, ignored. Worse, without that sort of methodology scrutiny, how many of the polls can become skewed?
Just like OJ and the press, the networks and newspapers have a symbiotic need for each other in an attempt to convince all of us, they are credible and trustworthy. It is nothing more than a feeble attempt to remain relevant in an era of media disintermediation. At the same time, each of us wants to feel our opinions are relevant as well.
Is it any surprise that we see the endless shifting back and forth between candidates in political polling.
I heard an interesting, if not impossible suggestion from a listener on KGO radio recently. The caller suggested that all presidential political ads be scrapped. Instead, each candidate would be given 60-90 minutes across all the networks to present a business plan for
their administration. The plan would have to be in specific detail.
Just as each of know from experience in business, a case would have to be made for the plan with expense and revenue justifications and substantiations. The business plans would be published for anyone to read.
I wonder if a poll were taken, asking us if we'd like to see a replacement of presidential campaign advertising and polling, with a program like the caller on KGO suggested, what the response would be."
Media Curmudgeon at January 15, 2008 06:30 PM writes:
Marilyn Keenan writes:
"I've been worried about these new voting machines and their software since hearing a story on the radio a couple years ago (probably NPR) about a test done here in Maryland. I'll forget the specifics, but basically the state hired a guy to come in and test whether or not the machines and their software could be circumvented.
He said he hacked into the state's vote count following the primary and "officially" changed the results. He said on the air that he did it without being detected and that it took only a few minutes. If he hadn't been doing it officially, he would have gotten away with it. So while all these discussions out there about machine fraud sound like "conspiracy theories", I think we all know that it is possible and that there are persons and groups out there perfectly willing to do whatever it takes to win. The question is whether or not the supposed safeguards in place really work. I'm all for a paper trail. There's too much at stake for cheaters to be able to change election
outcomes."
Post a comment
Thanks for signing in, . Now you can comment. (sign out)
(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)Printer-Friendly