Beware of statistics. Be even more wary of headlines.

First of all let’s talk about headlines. Good ones will grab your eye and entice you to read the story, or at least the first paragraph. If the first paragraph, the lead, is strong, then you may continue reading.

Online, the headline and lead sentence are even more important as your eyes dart all over the screen. That headline and lead may be competing with some distracting and gyrating animation. Argh!

When that headline and/or lead sentence disappoint — 'Click' — you move on, and the gyrating animation didn’t matter either.

So, on Sunday, July 31, 2011, the conservative-leaning Rasmussen poll sent an email with the headline: “30% Say Obama Too Confrontational, Highest Since Health Care Debate”. Thirty-percent of whom? Really? What does this mean? Is this factoid a big deal?

No, not really, I found after clicking through.

Thirty-percent isn’t many. Turns out, Rasmussen “buried the lead”. Deeper in the narrative, you’ll find the more salient information, if you make it that far.

“Forty-two percent (42%) of voters now view the president as a good or excellent leader, while 41% rate his leadership as poor, the latter his worst showing this year.”

“When he first took office, 64% gave him good or excellent marks for leadership. Those positives fell into the 40s by July 2009 and have generally remained in that range ever since.”

(Note: Gallup.com agrees: “Forty percent of Americans approve of the job Barack Obama is doing as president in Gallup's three-day rolling average for July 26-28, a new low for him by one percentage point.”)

I don’t know about you, but I think the two Rasmussen factoids found lower in the story, are more informative than their lead.

I should say that the question asked in the telephone survey of one thousand probable voters was unbiased and fair in my estimation: “How would you rate Barack Obama as a leader…excellent good fair or poor?”

Finally, Rasmussen’s survey methodology nowadays is becoming more and more suspect; they are calling landline telephones. Who is the most likely to have a hard-wired telephone, as more and more people are relying on their cellular telephones as their only phone?

According to a federal study by the Center for Disease Control and Prevention and reported by CBS News, in April, 2010 more than one in four households were cellphone-only — no landline. No doubt there are more now.

Here are the key questions: Who are these cell phone-only users? "Forty-four percent of folks between the age of 18 and 30 are wireless only. That's a huge number. And a lot of young adults will never have a landline in their lives,” the CBS report said. I used the CBS story, because the CDC study was dense and boring. (BTW: Looking at Texas, 32.5% of households were cell phone only in 2010.)

So, what demographic group had a landline last year? “Age is a big factor. Especially in the Northeast, where folks are older, they're less likely to have the cell phone,” said CBS.

Back to the results of the Rasmussen survey (which has a +/– 3% margin of error). You might expect the survey to be valid even if the headline and the lead were poorly written.

Let’s look at one more question: Who believes what?

“While 53% of Republicans believe Obama is too confrontational, 58% of Democrats say his leadership style is about right. Voters not affiliated with either political party are almost evenly divided on the president’s leadership style."

“Most GOP voters (73%) rate Obama’s leadership as poor, but 82% of Democrats feel he is doing a good or excellent job.”

The bottom line: This survey—it’s headline, its lead, and its methodology are at best all over the place in its present form, and at worst misleading. It shouldn’t be used by anybody to prove a political point.

Rasmussen needs to develop new methodology and hire new writers.

© Jim McNabb, 2011