I try not to respond to most inane ramblings by Chris Fitzsimon over at the NC Justice Center, but since he’s decided to distort the truth and omit critical pieces of information about our polling, I figured it was time.
Fitzsimon has been a bit flustered the past few months that we at Civitas do not send out press releases on every single poll question we ask. Considering that our polling is 35-40 questions in length each month, the mountains of press releases we’d have to produce would be unrealistic. Instead, we choose to highlight the key results in the poll through individual releases and then publish the entire results of the poll a few weeks later. Apparently that’s not good enough for him.
Now, this week Fitzsimon attempts to compare Obama’s favorability numbers from our May poll to our June poll — saying that Obama’s number’s “jumped 15 points.” On the surface, that is true. But Fitzsimon is misleading the readers of his column into thinking the polls are the same, when in fact they were conducted by two different pollsters (May by National Research Group, June by Tel Opinion Research) using two completely different likely voter screens. Fitzsimon knows this because he reads our press releases thoroughly and it’s stated right there in each one who the pollster is and what the methodology for that poll is. So in this case, he’s not lying, he’s just intentionally omitting facts to attempt to discredit Civitas polling.
Second, in an attempt to completely grasp at the last straw of his dignity, Fitzsimon attempts to say that including “leaners” in head-to-head matchups is somehow inappropriate. Again, Fitzsimon knows better and knows that this is a common practice in polling to allow undecided voters to choose which way they are favoring.
At the end of the day, Fitzsimon and the lefties have no choice but to run down Civitas polling simply because they don’t like the results. In true Clintonesque fashion, he can’t win philosophical arguments so he turns to the politics of personal (or in this case, institutional) destruction. How sad.
Oh Hayesy
Thanks for your concern and sorry that you are sad. And don’t worry I still have quite a few straws of dignity left.
I am not suggesting that you should send out press releases about every question in advance of your monthly poll results.
I am suggesting that you currently only send out press releases on findings that support your philosophical agenda. I think your devoted followers might benefit from press releases on a greater variety of questions and findings.
You do publish all the findings but the media reports on the ones you hightlight with press releases. That’s why you send them out.
I think it’s worth noting for example that the majority of North Carolinians believe that “Government hiring new employees helps offset the job losses suffered in the private sector and puts people back to work.”
As to the Obama approval numbers, did you announce in your press releases that it was a different pollster in May? There’d be no reason to think that if you use the same pollster month after month that you would switch, so it’s a safe bet a lot of people didn’t notice.
As to the voter screen, here is what you reported it was for the May poll.
This poll of 600 likely general election voters in North Carolina was conducted May 20, 22-23 by National Research, Inc. of Holmdel, NJ. All respondents were part of a fully representative sample of registered voters in North Carolina. For purposes of this study, voters interviewed had to have voted in two of the past four general elections or were newly registered to vote since 2008. An additional screening question was asked to filter only those voters having some likelihood to vote in the upcoming 2010 election.
Here is the one for June.
This poll of 600 likely general election voters in North Carolina was conducted June 15-18, 2010 by Tel Opinion Research of Alexandria, Virginia. All respondents were part of a fully representative sample of registered voters in North Carolina. For purposes of this study, voters we interviewed had to have voted in either the 2004, 2006 or 2008 general elections or were newly registered voters since 2008.
They are different, but not much, both trying to target people likely to vote in 2010. If you believe the screen to be different enough to produce dramatically different results, like a 15-point improvement in Obama’s approval rating, don’t you think it would have helped people to explain why the results were so different?
I am sure you wouldn’t want people to think Obama’s popularity has improved that much.
“I am suggesting that you currently only send out press releases on findings that support your philosophical agenda. I think your devoted followers might benefit from press releases on a greater variety of questions and findings.”
What an instructive suggestion. I’m sure that if Policy Watch ever conducted a poll, they would never cherry-pick only the results that “support your philosophical agenda” to include in their press releases.
Oh wait, luckily we don’t have to speculate! Let’s look at a press release from the July 9, 2007 Carolina Issues Poll from N.C. Policy Watch:
http://ncpolicywatch.com/docs/pdfs/Carolina_Issues_Poll_Media_Release_3.pdf
Here is the top paragraph: “North Carolina voters strongly oppose a plan favored by state Senate leaders to cut income taxes on the 60,000 wealthiest North Carolinians, but support extending the quarter cent sales tax hike first adopted in 2001, according the latest edition of the Carolina Issues Poll released Monday by N.C. Policy Watch.”
Hmmm, I wonder, however, what the opinion of those surveyed was about the overall tax burden in NC. If only you had asked them, you could have included that in the press release! Let’s look at the survey itself:
http://ncpolicywatch.com/docs/pdfs/Carolina_Issues_Poll_070507.pdf
Will you look at that? You did ask about it, and it was QUESTION #1 IN THE SURVEY!
But nowhere in the press release do you even mention the results of the FIRST QUESTION in the survey. Why did you not include the result of QUESTION #1 in the survey? It couldn’t have anything to do with the fact that the results showed that 64% said taxes were too high, could it?
Nah, couldn’t be. Because the fine folks at Policy Watch would never only include in their press releases survey results that fit their philosophical agenda (such as soaking the “rich”) while ignoring others (such as a majority of North Carolinians believing taxes are too high).
You’ve got quite a shiny glass house, mister stone-thrower.
And you guys accuse me of being flustered. Nice to see there’s still plenty of caffeine over there at Civitas.
You are welcome to point out that we didn’t issue a press release about a question in 2007 and we will continue to point out that you don’t ever issue them, month after month, when the findings aren’t exactly as you’d hoped.
I’d still be interested in the question of the 15 point improvement in Obama’s favorability index from May to June when the methodology seems very similar and the question was identical.
You can even reply in all caps if you’d like.
The frequency of the behavior has little relevance. The “but we only did it once” excuse is quite lame.
If I cheated on my wife once, I’d still be a hypocrite for condemming Tiger Woods.
Chris, it’s funny you point out Obama’s “15 point rise” from May to June, but neglect to point out that was preceded by a 10 point drop from April to May. Where’s that in your critique?
But since you asked about methodology here goes:
The April and June (and most other month’s) polls done by Tel Opinion are of people who voted in 1 out of the last 3 elections plus people who have registered since the 2008 election. Then they are asked if they are likely to vote, if they say they are “definitely not voting” the call is terminated.
In the National Research, Inc. (NRI)poll in May, you had to have voted in 2 out of past 3 elections and then were asked on a scale of 1-10 how likely you were to vote in 2010. If you said <4, the call was terminated.
NRI's "likely voter" screen was much, much tighter and not similar to that of Tel Opinion, thus I didn't state that Obama's numbers "moved" each month since the polling methodology wasn't comparable.
For the record, while NRI's likely voter screen seems tight, it was the exact one that they used to correctly predict the NJ Governor's race last year.