Sunday, August 27, 2006

That High Pitched Whine is the Media Spinning Their Poll Results

That high-pitched whine you hear is the establishment media covering their six on that Stephens-Media FOX 16 Oakleaf poll which claimed that Asa Hutchinson is hopelessly behind in the Governors race and the down ticket races are down hopelessly +. Brummett did a bit on it. The Arktimes burbled about it some. A normally clear Rob Moritz wrote (under pressure maybe?) a long muddled story that threw enough unrelated stats out to give the general impression that the numbers could be OK.

What is the reason they are circling the wagons like this? Simple. The numbers are not OK. The numbers are trash. The poll is flawed. This is not to make a disparaging personal attack on the Oakleafs. For all I know they are fine people. But anyone can mess up. They messed up and we explained how last week. It seems the Establishment media folks are getting some heat on this one, and that is why so much ink went out defending their poll. They are going to try to bluff their way through this, but they are still wrong.

This is the old establishment media trick where the candidate they disfavor is way down in a poll they commission in August. That crushes their momentum, confidence and fundraising. It makes them seem like "losers" in the eyes of the general public. Then, as it gets closer to election day, they issue polls that shows that the race is "narrowing", but alas, too late to help the disfavored candidate whose campaign had the wind knocked out of it's sails in the Summer.

They wind up losing in a fairly close race. A race they could have won if the public had known that the race was close all along. Do you realize that in 04 these same people had Jim Holt at around 30% right up until a week before the election? On election day he garnered more than 44% of the vote against Senator Blanche Lincoln. Now the same folks are trying to tell us that Holt has only 27% against some guy (Halter) who has lived out of state for the last 25 years, has never won an election in his life, and has managed to make some serious enemies in his own party during the short time he has been here.

Hey, we may be just a blog and they are the Establishment Media, but have we ever told you something that wrong before? Didn't think so. So believe us now when we assure you that they are wrong here. I am going to focus on the Moritz article, purely because it gave the most "facts" - as irrelevent and disassociated as they were- to support the Oakleaf poll.

(continued- click SUNDAY below and scroll down for the deconstruction of the Moritz article, or if sent straight here, just scroll down)

4 Comments:

Blogger Mark Moore (Moderator) said...

Mr. Moritz does have a good opening because he covers the controversy. The poll results "raise questions about the process". He quotes a good point from pollster John Zogby to the effect that the poll may simply be an outlier. That is to say, the poll has a 95% confidence level- but this may be the one time in twenty (5%) that the poll is wrong.

But that would imply their sampling methods were fine, they just got an odd group of respondents. The fact is, their sampling methods are skewed because the grossly over sample older voters who tend to vote democrat and grossly undersample the most conservative voting group- 18-34 year olds.

The Mortiz article gets close, oh so close, when they note that the poll did undersample younger voters, but makes no mention of the fact that they tend to vote Republican. They spend a lot of space talking about the switch to cell phones as a possible reason this group is undersampled, but NO space explaining how that could skew the overall results.

Only near the end was the biggest sampling problem mentioned- gigantic over representing of older voters- but mostly in an effort to explain it away. Once again, no mention was made of how this would skew the results of the poll. Older voters tend to vote democratic, even if they agree with the Republican on all the issues, but the article did not mention this.

He writes, "The percentage of older respondents in the Opinion Research poll was not much different than the percentage of older voters who cast ballots in the 2004 presidential election in Arkansas, according to data collected by the Republican Party of Arkansas.

Party workers canvassed the state on Election Day in 2004 and interviewed 1,459 people as the(y) exited the polls. Their results showed 55 percent of respondents were 45 or older. Twenty-nine percent were between 45-59 and 26 percent were 60 or older."

A lot of facts there, but not relevant facts. It is the 65+ crowd that gives the democrats the biggest margin. Mortiz fails to reveal that the Oakleaf poll had a whopping 36% from 65+ voters! Instead, he clouds it by grouping it with younger voters, 55 and up or even 45 and up. But the plain fact is that of actual voters on election day in 04, only 26% were 60 or older while in the Oakleaf poll 36% were 65 and older.

So then how many voters on election day in 04 were 65 or older? Most people don't last much past 75, and I would guess most who do are interested in eternity more than voting. A reasonable guess is that a quarter to a third of all voters over 60 are between 60 and 65. My wife the election worker confirms this estimate. So say that means that you can shave about a third off of the numbers in the 04 post election poll to get an estimate of the percentage over 65. In other words, if 26% were over 60 then 26% X 0.66 were over 65.

That is just a hair over 17% in actual voters being over 65 vs. 36% of voters over 65 in the Oakleaf poll. They need to give it up!

The poll under-represented the age group that votes most one way and over represented the group that votes most the other way. That is why their numbers are wrong, and I don't know why they won't just say that. Its blindingly obvious.

The last couple of paragraphs note that questioning polls are common "particularly among those who don't like the results."

Please. We have every right to complain about this snow-job. Even if the papers call a cow-patty a rose, it is still going to draw flies. You can blame that on the flies being wrong about the smell, or you can blame it on the press being wrong about the rose.

6:22 PM, August 27, 2006  
Anonymous terrymcdermott said...

Most polls will give Democrates a 5% advantage. Just give that 5% to the Republican and you have a more realistic poll. And when there is third party or independent you take away another two points from the Democrate.

6:23 PM, August 27, 2006  
Anonymous Anonymous said...

Thank you Mark for this great analysis. My question is do you think the MSM ever feels guilty for the way they treat conservatives, the way they manipulate, show favoritism.

I get so frustrated - it isn't fair and they seem to relish their role in distorting the truth.

7:05 PM, August 27, 2006  
Blogger Mark Moore (Moderator) said...

It is a process. Some do feel guilty, but it just takes time and understanding. Most go in with the idea that we are some kind of monsters or fanatics and they are serving the "greater good" by keeping us down. Others honestly can't see that they are biased, to them it is just the way "enlightened people think".

With time we can show them that we have a position, and a valid one at that.

Others don't feel guilty and get off on manipulating the public. Those are the ones I feel sorry for- eternally speaking. The ones who know what they are doing and do it anyway. The ones who come to awareness, but it does not matter.

We need to reach out to them too- to smack them in the mouth. Well that was a little joke there, but it really is sad.

8:25 PM, August 27, 2006  

Post a Comment

Links to this post:

Create a Link

<< Home