Matt Towery's Inside The Numbers:

CBS Obama Poll Misleading, Flawed And Biased

By Matthew Towery

(3/25/08) Last week I wrote that it would be hard to determine the political impact of either Barack Obama's pastor's comments, or Obama's subsequent speech about those various comments. I also wrote that our own polling firm would attempt to measure exactly that.

(For the record, our polls have correctly polled the winner in every presidential primary race we've surveyed this year, except the Tennessee Republican primary.)

In our survey of the Wright / Obama situation, we merely asked if respondents were aware of the pastor's "past comments," and of "Sen. Obama's speech about the remarks made by his pastor." We did not ask respondents if they heard Obama's speech, or knew the (racial) nature of the pastor's comments over the years.

The results indicated not only that a significant percentage of white voters, and independent voters of all ethnicities, were now "overall" less likely to vote for Obama for president, but also that a plurality of African-Americans felt the same way.

Admittedly, our survey was only a snapshot of public opinion. And its results likely were based on vague notions by respondents of the Rev. Jeremiah Wright's comments and of Obama's speech. I would never claim the poll as definitive, or as an indication of what influence the whole affair may eventually have on the presidential race.

Then I came across what in my view is the single most biased and dishonest public opinion survey I've ever seen.

CBS News released a poll that read, "Most voters following the events regarding Senator (STET) Barack Obama and Rev. (STET) Jeremiah Wright think Obama's speech was a success."

This lead sentence in the press release appeared in media outlets all over the world. It gave the impression that Obama's speech on Wright had ended any dispute or controversy about Obama being associated with Wright and his many inflammatory comments.

Maybe over time that will prove true. And Obama's speech was an impressive one, no doubt.

But that's not the point. Most Americans didn't hear the speech. Most have only heard bits and pieces about this whole business about Obama and his pastor of many years.

So, rather than survey registered voters in the customary way -- by randomly finding voters around the nation and asking them their impressions of the issue -- CBS instead adopted the most curious polling methodology I've ever some across. It was probably the only methodology possible on this subject that could reach what was apparently the network's desired outcome.

CBS didn't randomly phone registered voters. They didn't weight the poll for age, race, gender and political affiliation. Instead, they chose to poll a group of people prior to Obama's speech about Wright. The poll asked respondents their opinions about Wright and his views.

Then CBS later re-polled the exact same group to gauge their reactions to Obama's speech. Based on those two separate pollings, the network extrapolated the result that "Sixty-nine percent of voters who have heard or read about Obama's speech think he did a good job addressing the issue of race relations."

Unbelievable! Had my firm employed these types of polling tactics, pundits and alleged "polling experts" would have torn us to pieces.

To fully understand how CBS's methodology here was biased and flawed, let's use common sense.

First, you get a call from CBS pollsters, prior to the Obama speech, asking you what you think about Wright and those of his past views that have made for controversy in the news lately. You answer the survey.

Now you are keenly aware of the issue, because you have just been polled about it, probably in some detail.

Would it not follow that you are more likely now to pay close attention to the issue? To view the speech, or read it word for word? Of course.

Can we really believe that the people at CBS, who are obviously bright, didn't understand the likely impact that the results of their first poll would have on the second poll of the exact same people? That's not a random poll. It's a giant focus group.

For CBS to represent that calling the same group twice on the same issue is a reflection of American public sentiment is a farce. This so-called poll was the very sort of fast and loose twist on legitimate polling and journalism that has so many Americans mistrustful of national media. This was nothing less than a case of network-distributed damage control on behalf of Barack Obama. It's that simple.

As for our poll of the same situation: At least I was fair enough to admit that I can't yet say that Obama's speech has severely damaged his ultimate chances of becoming president. That, even though our survey showed many respondents indicating they were less likely to vote for Obama after the Wright comments and the Obama speech. And I don't have to survey the same people I surveyed last week to prove my own theory.

--------------------------------------------------------------------------------

Matt Towery served as the chairman of former Speaker Newt Gingrich's political organization from 1992 until Gingrich left Congress. He is a former Georgia state representative, the author of several books and currently heads the polling and political information firm InsiderAdvantage. To find out more about Matthew Towery and read features by other Creators Syndicate writers and cartoonists, visit the Creators Syndicate website at www.creators.com.

COPYRIGHT 2007 CREATORS SYNDICATE, INC.