NYT/CBS poll has its thumbs on the scale … again

posted at 12:21 pm on April 7, 2009 by Ed Morrissey

The only consistent quality about the New York Times/CBS polling is its ridiculous sampling on party identification.  Once again, the NYT/CBS poll tries to paint Democrats as unassailably popular and Republicans as losing ground, and once again they have to wildly oversample Democrats to make the argument.  Their latest poll has a sixteen-point gap between the parties:

Americans have grown more optimistic about the economy and the direction of the country in the 11 weeks since President Obama was inaugurated, suggesting that he is enjoying some success in his critical task of rebuilding the nation’s confidence, according to the latest New York Times/CBS News poll.

These sometimes turbulent weeks — marked by new initiatives by Mr. Obama, attacks by Republicans and more than a few missteps by the White House — do not appear to have hurt the president. Americans said they approved of Mr. Obama’s handling of the economy, foreign policy, Iraq and Afghanistan; fully two-thirds said they approved of his overall job performance.

By contrast, just 31 percent of respondents said they had a favorable view of the Republican Party, the lowest in the 25 years the question has been asked in New York Times/CBS News polls.

It is not unusual for new presidents to enjoy a period of public support. Still, the durability of Mr. Obama’s support contrasts with that of some of his predecessors at the same point in their terms. It is also striking at a time when anxiety has gripped households across the country and Mr. Obama has alternately sought to rally Americans’ spirits and warn against economic collapse as he seeks Congressional support for his programs.

It’s not unusual for the Gray Lady to cook the numbers, either, to make sure their poll shows that support.  This is the breakout in their demographics on page 23:

  • Democrats – 39%
  • Republicans – 23%
  • Independents – 30%

In February, just a month earlier, they had Democrats at 36% and Republicans at 26% — still too low for the GOP, but only a ten-point gap.  Now they want to argue that Democrats have pushed the partisan gap to 16 points in a single month?  Not hardly, says Rasmussen:

In March, the number of Democrats in the nation fell two percentage points while the number of Republicans fell by half-a-point. Democrats continue to have a sizable advantage in terms of partisan identification, but the advantage is smaller than it’s been since December 2007.

Currently, 38.7% of Americans say they are Democrats. That’s down from 40.8% a month ago. It’s also the first time the Democratic total has slipped below the 40% mark since the Republican convention bounce last September. … In March, 33.2% of American adults say they are Republicans. That’s down from 33.6% a month ago. Over the past year, the number of Republicans has ranged from a low of 31.4% in April to a high of 34.4% in September.

In other words, the percentage has never been as low as 23%.  Republicans lost the national election, but if they had dropped to 23% nationwide, Democrats would likely control over 300 seats in Congress and 70 seats in the Senate.  Any poll using a 16-point gap between the parties is worthless.

Unfortunately, they’re not the only news agencies using ridiculous sampling to bolster their contentions of wild popularity for Barack Obama.  Newsweek used a 19-point gap in its latest polling.  CNN thus far has neglected to publish the internals of its poll.  It’s just another round of dishonesty from the media.

Rasmussen has Obama’s polling at a more rational 58%, but then again, they don’t have an investment in Obama, Inc, as the media does.

Update: Jim Geraghty reminds us that in a highly successful electoral cycle, Democrats had an affiliation edge of …. seven percent.

Blowback

Note from Hot Air management: This section is for comments from Hot Air's community of registered readers. Please don't assume that Hot Air management agrees with or otherwise endorses any particular comment just because we let it stand. A reminder: Anyone who fails to comply with our terms of use may lose their posting privilege.

Trackbacks/Pings

Trackback URL

Comments

Comment pages: 1 2 3

They’re getting skewed samples in order to deliver results they like. And they get away it because their apologists don’t demand honest polling — including you.

See, Ed, this is why you’d never make it as a reporter. You can’t just make stuff up and report it without evidence to back up your claim.

You have NO — as in zero, zilch — evidence that the NYTimes purposefully sought a sample with a larger amount of democrats than republicans so they could provide positive numbers for Obama. And the evidence — a larger amount of people IDing themselves as Dems across the nation; the fact that more of the people polled IDed themselves as conservative rather than liberal — points to a random sampling of people, more of which ID themselves as democrats. This isn’t a random poll of Republicans and Democrats, but a random poll of Americans.

You make the charge that the NYTimes is purposefully polling more Dems with zero evidence to support it — and despite evidence to the contrary.

That is dishonest of you.

Tom_Shipley on April 7, 2009 at 1:28 PM

NYT/CBSRassmussen poll has its thumbs on the scale … again

Corrected.

Rassumssen on the other hand has been asking questions in some very slanted ways that skew GOP. Thankfully we have plenty of pollsters to choose from that are more reliable, including even the NYT/CBS poll, believe it or not.

No wonder Fox quotes Rassmussen.

sethstorm on April 7, 2009 at 1:30 PM

Sorry for the double post… didn’t think the first one went through.

…they poll people at random and only find out the breakdown of party affiliation after the poll is taken.
Tom_Shipley on April 7, 2009 at 12:52 PM

Do you really believe this?

Well, considering one of the final questions asks those polled to ID their party affiliation, and there’s no evidence to show that the Times did anything but a random sample, yes.

Tom_Shipley on April 7, 2009 at 1:33 PM

Pollsters they have discovered over time that they can drive public opinion by these skewed polls. Nothing in media can be trusted anymore. Our only real hope is to expose all of the liars at all media outlets.

What a world.

petunia on April 7, 2009 at 1:33 PM

Another reason to believe nothing from the established propaganda machine.

Mason on April 7, 2009 at 1:35 PM

You make the charge that the NYTimes is purposefully polling more Dems with zero evidence to support it — and despite evidence to the contrary.

That is dishonest of you.

Tom_Shipley on April 7, 2009 at 1:28 PM

Seems to me the evidence is in the numbers. If the times had culled the Democrat responses to be in line with national figures then the results would be more palatable.

oldernwiser on April 7, 2009 at 1:36 PM

Independents – 30%

Half the “independents” I’ve met were raving lunatics. They are usually worse than registered demofascists.

peacenprosperity on April 7, 2009 at 1:37 PM

You make the charge that the NYTimes is purposefully polling more Dems with zero evidence to support it — and despite evidence to the contrary.

That is dishonest of you.

Tom_Shipley on April 7, 2009 at 1:28 PM

Did you not read my post? Look at the actual pdf if you don’t believe me. They reweighted the numbers to increase democrats and decrease republicans.

Vashta.Nerada on April 7, 2009 at 1:37 PM

I suppose all these jackass libs saw these BS poll numbers and figured they’d come here to somehow gloat. When we have double-digit unemployment a year from now and the media gives him these kinds of unreal poll ratings are you idiots gonna finally get a clue.

No president has these kinds of poll ratings unless the economy is doing well(it’s not) or if there is some kind of national emergency(9/11) or a hot war that is getting nonstop attention(Gulf War).

gsherin on April 7, 2009 at 1:38 PM

Such are the actions of the scared ObamaMedia puppets.

DannoJyd on April 7, 2009 at 1:38 PM

You make the charge that the NYTimes is purposefully polling more Dems with zero evidence to support it — and despite evidence to the contrary.

That is dishonest of you.

Tom_Shipley on April 7, 2009 at 1:28 PM

Yes, and you are outrageously outraged. We get it. Evidence to the contrary? He cites the samples of another poll, one which was far more accurate in predicting the actual outcome of the actual election.

You could perhaps argue this as a case of unintentionally flawed methodology, but not the fact that the NYT/CBS poll has been wildly inaccurate, and in the Democrats’ favor.

A Balrog of Morgoth on April 7, 2009 at 1:38 PM

You have NO — as in zero, zilch — evidence that the NYTimes purposefully sought a sample with a larger amount of democrats than republicans so they could provide positive numbers for Obama.
That is dishonest of you.

Tom_Shipley on April 7, 2009 at 1:28 PM

I won’t go so far as to say you are being dishonest here, I will stop at woefully uninformed. I have an undergrad minor in stat and plenty of grad exposure. The population is determined at the design of the study. Have you not had the experience where being called for a study you met a criteria that was already filled in sampling? To say this is not predetermined shows no understanding of the research design process.

msmveritas on April 7, 2009 at 1:40 PM

Did you not read my post? Look at the actual pdf if you don’t believe me. They reweighted the numbers to increase democrats and decrease republicans.

How did they “reweight” the numbers?

Tom_Shipley on April 7, 2009 at 1:40 PM

Since they show how they changed the party affiliation weights, wouldn’t it be simple math to discover what their results would be if they just stuck with their original numbers, or if they used Rasmussen’s numbers?

The weighted numbers:
Dem: 39%
Rep: 23%
Ind: 38%

The unweighted numbers:
Dem: 35%
Rep: 26%
Ind: 39%

Overall Approve / Disapprove / Not Sure
Dem: 89% / 7% / 4%
Rep: 31% / 54% / 15%
Ind: 63% / 24% / 13%

Weighted Total: 66% / 24% / 10%
Unweighted Total: 64% / 26% / 10%

Now, finally, with Rasmussen’s numbers:
Dem: 38.7% / Rep: 33.2% / Ind: 28.1%

Dem: 386 / Rep: 331 / Ind: 280

Dem: 344 / 27 / 15
Rep: 103 / 179 / 50
Ind: 176 / 67 / 36
Tot: 623 / 273 / 101

Rasmussen Weights: 62.4% / 27.4% / 10.2%

Anyone is free to check my math, if bored.

strictnein on April 7, 2009 at 1:41 PM

Rassumssen on the other hand has been asking questions in some very slanted ways that skew GOP. Thankfully we have plenty of pollsters to choose from that are more reliable, including even the NYT/CBS poll, believe it or not.
No wonder Fox quotes Rassmussen.

sethstorm on April 7, 2009 at 1:30 PM

Right. Reliable. Just ignore how well a particular polling firm did in predicting the election results with the actual election results.

I guess when you went shopping for your car, a “reliable” source of replacement parts was a high priority.

Idiot. At least Shipley is making an argument that deserves a more thoughtful response than GFY.

A Balrog of Morgoth on April 7, 2009 at 1:44 PM

sethstorm on April 7, 2009 at 1:30 PM

Chew on this…and don’t let the facts crash your hard drive:

Who Were The Most Accurate Presidential Pollsters?
Written by: Evrviglnt on Wednesday, November 5th, 2008

A Fordham University report by Costas Panagopoulos rates the pollsters for this presidential election. Here are the most accurate:

1T. Rasmussen (11/1-3)**
1T. Pew (10/29-11/1)**
3. YouGov/Polimetrix (10/18-11/1)
4. Harris Interactive (10/20-27)
5. GWU (Lake/Tarrance) (11/2-3)*

6T. Diageo/Hotline (10/31-11/2)*
6T. ARG (10/25-27)*
8T. CNN (10/30-11/1)
8T. Ipsos/McClatchy (10/30-11/1)
10. DailyKos.com (D)/Research 2000 (11/1-3)

11. AP/Yahoo/KN (10/17-27)
12. Democracy Corps (D) (10/30-11/2)
13. FOX (11/1-2)
14. Economist/YouGov (10/25-27)
15. IBD/TIPP (11/1-3)

16. NBC/WSJ (11/1-2)
17. ABC/Post (10/30-11/2)
18. Marist College (11/3)
19. CBS (10/31-11/2)

20. Gallup (10/31-11/2)
21. Reuters/ C-SPAN/ Zogby (10/31-11/3)
22. CBS/Times (10/25-29)
23. Newsweek (10/22-23)

Link

RepubChica on April 7, 2009 at 1:45 PM

Links/etc of any type to show how his polling is partisan would be great to see.
http://www.pollster.com/blogs/how_pollsters_affect_poll_resu.php

sesquipedalian on April 7, 2009 at 12:41 PM

You know there is a problem with what that site tries to accompish. Zero is the actual voting data for the country. Rasmussen nailed the voting data when it was all said and done yet its “house effect” is listed as stronger for McCain on that site.

Polling house effect should consider accuracy as the foremost concern over push for response or demographic leanings. Short version : if you take an average of bad data you wind up with a bad average, you dont wind up with accuracy.

Rasmussen was accurate, the others were not. Maybe Rasmussen isnt right leaning, maybe they are just right.

Opportunity Costs on April 7, 2009 at 1:45 PM

How did they “reweight” the numbers?

Tom_Shipley on April 7, 2009 at 1:40 PM

Thanks, strictnein.

Vashta.Nerada on April 7, 2009 at 1:45 PM

Again, how were the numbers reweighted? Was it done for other reasons, such as over-representation of one area of the nation over others? Which dovetails nicely into this latest polling goodness from the Great Orange Satan hisself:

President Obama favorability:

Northeast: 88% favorable, 11% unfavorable
Midwest: 73% favorable, 24% unfavorable
West: 76% favorable, 22% unfavorable
South: 41% favorable, 54% unfavorable

Democratic Party:

Northeast: 66% favorable, 23% unfavorable
Midwest: 52% favorable, 39% unfavorable
West: 54% favorable, 38% unfavorable
South: 31% favorable, 61% unfavorable

Republican Party:

Northeast: 8% favorable, 82% unfavorable
Midwest: 22% favorable, 68% unfavorable
West: 20% favorable, 70% unfavorable
South: 43% favorable, 47% unfavorable

What’s that about being a “national party no more”, Zell? Zell??

starfleet_dude on April 7, 2009 at 1:46 PM

Even Kos’ polls faired better than CBS/NYT during the presidential election–very telling indeed….

RepubChica on April 7, 2009 at 1:46 PM

Thankfully we have plenty of pollsters to choose from that are more reliable, including even the NYT/CBS poll, believe it or not.

……

What evidence do you have they are more reliable?

artist on April 7, 2009 at 1:47 PM

The population is determined at the design of the study. Have you not had the experience where being called for a study you met a criteria that was already filled in sampling? To say this is not predetermined shows no understanding of the research design process.

I will claim ignorance about how actual sampling occurs. But, I still see no evidence that the Times/CBS poll knows the party affiliation of those sampled prior to polling them.

Tom_Shipley on April 7, 2009 at 1:47 PM

RAS nailed the 08 election, CBS/NYT di not.

That is a FACT.

artist on April 7, 2009 at 1:48 PM

Since they show how they changed the party affiliation weights,

What page of the PDF is this on?

Tom_Shipley on April 7, 2009 at 1:49 PM

starfleet_dude on April 7, 2009 at 1:46 PM

For those of us unfamiliar with the color coded satans, how about a link?

strictnein on April 7, 2009 at 1:52 PM

What page of the PDF is this on?

Tom_Shipley on April 7, 2009 at 1:49 PM

Last page.

UNWEIGHTED WEIGHTED
Total Respondents 998
Total Republicans 264 230
Total Democrats 349 391
Total Independents 385 377
Half Sample A 485 483
Half Sample B 513 515

strictnein on April 7, 2009 at 1:53 PM

Tom_Shipley on April 7, 2009 at 1:47 PM

They do not know but in a reliable study there would be a quota of x Reps, x Dems and x Independents. When calling if the quota for Dems is met, each additional call made to a self identified Dem voter should put that call into an opt out and the rest of the study with that participant would not be taken. You want the study to be recreatable with the entire population and to do that you need a confidence level, meaning if you called every single person in the US of voting age your results would be roughly the same. A skewed sample results in a skewed result, its research design 101.

msmveritas on April 7, 2009 at 1:54 PM

See, Ed, this is why you’d never make it as a reporter. You can’t just make stuff up and report it without evidence to back up your claim.

Tom_Shipley on April 7, 2009 at 1:28 PM

Earth to Shipley-”reporters” do that all the time. Ask Dan Rather.

Del Dolemonte on April 7, 2009 at 1:56 PM

See, Ed, this is why you’d never make it as a reporter. You can’t just make stuff up…

Tom_Shipley on April 7, 2009 at 1:28 PM

ROFL!!! Should have stopped when you were ahead. Ed has too much honesty and integrity to be a reporter. But if he wanted a dead-end career at a dead-end newspaper, he could always learn how to make stuff up…

dominigan on April 7, 2009 at 1:59 PM

Rassumssen on the other hand has been asking questions in some very slanted ways that skew GOP. Thankfully we have plenty of pollsters to choose from that are more reliable, including even the NYT/CBS poll, believe it or not.

starfleet_dude on April 7, 2009 at 1:24 PM

Ever been polled by Scott’s firm? I have.

Very professionally done, no biased questions.

Del Dolemonte on April 7, 2009 at 2:00 PM

try this http://www.cbsnews.com/htdocs/pdf/poll_Obama_040609.pdf

All the way at the bottom. Page 24.

Opportunity Costs on April 7, 2009 at 2:00 PM

ROFL!!! Should have stopped when you were ahead. Ed has too much honesty and integrity to be a reporter. But if he wanted a dead-end career at a dead-end newspaper, he could always learn how to make stuff up…

dominigan on April 7, 2009 at 1:59 PM

Props to you–nicely done!

RepubChica on April 7, 2009 at 2:03 PM

What page of the PDF is this on?

Tom_Shipley on April 7, 2009 at 1:49 PM

Last page.

UNWEIGHTED WEIGHTED
Total Respondents 998
Total Republicans 264 230
Total Democrats 349 391
Total Independents 385 377
Half Sample A 485 483
Half Sample B 513 515

strictnein on April 7, 2009 at 1:53 PM

“We pause now for station identification while Tom_Shipley tries to figure out how to get his foot out of his mouth, and his head out of his @$$…

You have just been reading HOTAIR.com“… doh! doh! DOH!

dominigan on April 7, 2009 at 2:04 PM

They not only lie about the numbers they are lying about their assertion that most other presidents have not retained this sort of popularity this far into their presidency. In fact most presidents are still quite popular a mere 100 days after entering office and Obama is not there yet.

Pretty soon if his numbers start to slip they will be announcing there are no longer any Republicans in the United States and thus will poll only Democrats and Democrat leaning Independents. Kind of like Saddam’s numbers.

Terrye on April 7, 2009 at 2:06 PM

Terrye on April 7, 2009 at 2:06 PM

You know, I find it amusing watching them bend over backwards, distorting their numbers enough to get the results they desire.

They must really be disparate!

dominigan on April 7, 2009 at 2:09 PM

Another thought: These types of approve/disapprove polls can hide a lot of movement. A poll with varying levels of approval and disapproval would be much more interesting.

66% approval could be 42% Strong Approval and 24% Moderate Approval, or 24% Strong Approval and 42% Moderate Approval.

strictnein on April 7, 2009 at 2:09 PM

starfleet:

Oh yeah, Rasmussen is skewered, like do you want to pay more taxes? vs Do you want your children to receive health care or would you prefer they die a slow and ugly death from some other polling outfit.

Terrye on April 7, 2009 at 2:10 PM

Err…

They must really be desperate!

Sometimes not even spell-check is enough to save you…

dominigan on April 7, 2009 at 2:10 PM

dominigan on April 7, 2009 at 2:04 PM

OK, so let me ask you this: What the difference between the weighted and unweighted results? What’s done to the results to make them “weighted”?

Anyone know this?

When calling if the quota for Dems is met, each additional call made to a self identified Dem voter should put that call into an opt out and the rest of the study with that participant would not be taken.

So let me ask you this? What cap should be used when determining their quota for Dem voters v. Republican voters?

As I said before, this is a poll of Americans, not a poll of Democrats and Republicans. Since Dems have a higher % of party ID nationwide (12 points according to some polls), if the NYTimes samples people without taking party affiliation into consideration, is it that big a deal if there’s 16% gap in party affiliation?

And, again, I have yet to see ANY evidence that shows the NYTimes tried to skew the sample toward democrats.

Tom_Shipley on April 7, 2009 at 2:10 PM

dominigan:

Well there will be an election in 2010 and they have to start making people feel hopeless, out of step, and outnumbered now if they intend to dampen voter turnout in the midterms.

Terrye on April 7, 2009 at 2:11 PM

UNWEIGHTED WEIGHTED
Total Respondents 998
Total Republicans 264 230
Total Democrats 349 391
Total Independents 385 377
Half Sample A 485 483
Half Sample B 513 515

strictnein on April 7, 2009 at 1:53 PM

HTH
the total number of respondents is 998
after weighting the total of Democrats increases by 42
while the republicans and independants decrease by a combined 42. The decrease is easy to figure out, the Times (as is their wont) just ignores them.
But where do the extra 42 come from?
ACORN!
:-)

oldernwiser on April 7, 2009 at 2:12 PM

Tom:

The other day a friend of mine got a call from a polling agency, when ask if she was a Democrat or Republican or Independent, she responded Republican. That was the end of the conversation. It seems they were looking for Democrats to poll.

Terrye on April 7, 2009 at 2:13 PM

But where do the extra Democrat 42 come from?
fixed

oldernwiser on April 7, 2009 at 2:13 PM

Ever been polled by Scott’s firm? I have.

Very professionally done, no biased questions.

Del Dolemonte on April 7, 2009 at 2:00 PM

I’ve been polled by Zogby, and it was well done. Needless to say, the plural of anecdote isn’t data.

starfleet_dude on April 7, 2009 at 2:13 PM

Hey, but I could do my own little poll. The vast majority of the people I come into contact in my little part of the world are not Obamabots. Not at all. I would say that is every bit as reliable as CBS.

Terrye on April 7, 2009 at 2:14 PM

starfleet:

Rasmussens questions are no more skewered than any other polling agency. As has been noted, these are pubic opinion polls, and it is difficult to know for sure if they are accurate. One of the few ways to verify their findings is with elections…and Rasmussens has a very good track record when it comes to elections.

Terrye on April 7, 2009 at 2:16 PM

That is one thing I found interesting when I read Atlas Shrugged. It was written more than a half century ago and it shows the beginning of governing by polling and manipulating polls. The idea is the way to manipulate public opinion rolls is to manipulate public opinion.

Terrye on April 7, 2009 at 2:18 PM

This “poll” is nothing more than manufactured propaganda for the dumbMasses who have yet to catch on to what Little Lord Obamaroy is really up to.

The MSM will do whatever it takes to prop up the miserably failing potentate in the White House, and that includes blatantly lying through their teeth whenever they deem it necessary.

The Obama “presidency” is a rapidly unfolding large-scale disaster. Even the most hard-core, true believing among the freedom-hating lefties know this deep down. They just refuse to admit it.

We are in the very early stages of what will eventually turn out to be the most dishonest, corrupt, incompetent and dangerous administrations in the 236 year history of this nation.

-Dave

Dave R. on April 7, 2009 at 2:18 PM

Great numbers for the President.

getalife on April 7, 2009 at 2:20 PM

getalife on April 7, 2009 at 2:20 PM

Yeah and they’re about as authentic as the tooth fairy or Santa Claus.

gsherin on April 7, 2009 at 2:23 PM

I saw this poll earlier. I got about 3 lines into the article and saw the words “NY Times/CBS” and that told me all I needed to know.

Is Rasmussen’s poll showing Obama at 58% accurate? Who knows? What I do know is this. He nailed the 2008 Presidential election and he’s had Obama consistently around 56-58% for the better part of a month now. I’d say he deserves the benefit of the doubt a lot more than the nearly bankrupt New York Times or a network whose stock is at 4 bucks.

Doughboy on April 7, 2009 at 2:23 PM

The best resource for polling data is Real Clear Politics. Here is the link for Obama’s job approval rating — averaged between the various sources and with historical results and trends:

http://www.realclearpolitics.com/epolls/other/president_obama_job_approval-1044.html

cruadin on April 7, 2009 at 2:25 PM

getalife:

When Bush’s were 90%, did you think they were great? I bet you did. yeah sure.

Terrye on April 7, 2009 at 2:26 PM

I hope they fail.

bluelightbrigade on April 7, 2009 at 2:29 PM

The thing is CBS is the outlier here. Other polls are a lot closer to Rasmussens than they are to CBS.

Terrye on April 7, 2009 at 2:29 PM

*They = CBS/NYT

bluelightbrigade on April 7, 2009 at 2:29 PM

getalife on April 7, 2009 at 2:20 PM

Oooohhh scarrryyy…boogah-boogah!

bluelightbrigade on April 7, 2009 at 2:31 PM

The interesting thing is that as more and more polls show Obama in the 50′s, CBS comes out with this. It is almost as if they are trying to bring up his average. At this same time in his presidency Bush’s numbers were between 56 and 63. After three years he average 63%. So I guess CBS’s assertion that there is something special about Obama’s numbers is just bs. But then again this is CBS.

Terrye on April 7, 2009 at 2:31 PM

And, again, I have yet to see…

Tom_Shipley on April 7, 2009 at 2:10 PM

That would be much more honest than anything you have posted here today Tom…

Exit Question: Will Tom Shipley have the intellectual integrity to apologize for accusations made to Ed Morrisey?

Keemo on April 7, 2009 at 2:32 PM

Keemo on April 7, 2009 at 2:32 PM

I won’t apologize, because nothing I’ve said is wrong. Ed accused the NYTimes of “cooking the numbers” so it polled more Dems without any evidence of this. That’s dishonest. But, being a blogger, he can just post whatever opinion he has without backing it up with any fact.

Tom_Shipley on April 7, 2009 at 2:37 PM

I think there biggest problem is that is who listens/watches CBS news. If I went to ShawnHannity.com I would probably see a poll that has like 85% Republicans. No reporter would ever say that is a valid poll.

The issue is that CBS still decided to report a bad poll as valid.

jeffn21 on April 7, 2009 at 2:42 PM

once again they have to wildly oversample Democrats to make the argument

I keep sayin it. There are five ways to make a poll come out the way you want it….

dogsoldier on April 7, 2009 at 2:47 PM

Tom_Shipley on April 7, 2009 at 2:10 PM

I would have used the breakdown from the last election at a minimum. I believe that was 39 Dem, 32 Republican and 29 Independent. Now there is an argument to be made that the voting population is not necessarily the same as the population at large but it is something that I would say has a strong confidence level. That is the + or – level you say your results could be repeated in the perfect sample (everyone).

Still that is not a static number because recent polls show Dem voter ID down for the first time since 2004 by approx 2 points, Rep -.5 points and Independent up. If you followed the voting sample from 2008 you would likely still oversample Dems but it would be much more reliable that what we see in this poll. This poll is garbage, any statistician would tear this apart.

msmveritas on April 7, 2009 at 2:47 PM

If I went to ShawnHannity.com I would probably see a poll that has like 85% Republicans

No, you would not. Conservatives do not need to cheat. We’ll take the hit if there is one and stick to honest work. Gallup recently put out a poll showing Hussein at under 50%. I think its probably less than that.

dogsoldier on April 7, 2009 at 2:50 PM

See, Ed, this is why you’d never make it as a reporter…

Tom_Shipley on April 7, 2009 at 1:28 PM

Ed never said he was a reporter. He’s entirely upfront about his leanings. Put down the glue, Tom.

John the Libertarian on April 7, 2009 at 2:51 PM

Such are the actions of the scared [insert opponent here]

DannoJyd on April 7, 2009 at 1:38 PM

Is that correct?

sethstorm on April 7, 2009 at 2:55 PM

I won’t apologize, because nothing I’ve said is wrong. Ed accused the NYTimes of “cooking the numbers” so it polled more Dems without any evidence of this. That’s dishonest. But, being a blogger, he can just post whatever opinion he has without backing it up with any fact.

Tom_Shipley on April 7, 2009 at 2:37 PM

You wrote this before I explained what facts Ed bases his claim of “cooking the numbers” so I just want to reiterate, look at the update and notice the spread of the last election was 7 points, not 16. Polls indicate Dem ID going down not up. There is no logical argument that a 16 point spread represents any statistically valid finding. The study is completely worthless.

msmveritas on April 7, 2009 at 2:56 PM

Tom_Shipley on April 7, 2009 at 1:28 PM

Dude, it’s the New York Times with CBS.

It should be 100% ignored.

If it ain’t Rasmussen, it can be ignored. If it ain’t Rasmussen, it is s**t.

Sapwolf on April 7, 2009 at 2:57 PM

If it ain’t Rasmussen, it can be used. If it is Rasmussen, it is s**t meant for Fox’s use.

Sapwolf on April 7, 2009 at 2:57 PM

Corrected.

sethstorm on April 7, 2009 at 3:01 PM

Not space aliens, Lo…Chupacabras.

Mike D. on April 7, 2009 at 12:30 PM

Cut out that talk of illegals.

You are gonna spook my production line.

Sapwolf on April 7, 2009 at 3:01 PM

This is from Ed’s link from Geraghty, the election is November 2008:

Democrats’ Party ID advantage, according to exit polls: 39 percent Democrat, 32 percent Republican, 29 percent Independent.

It was on the far edge, but within the range I had been thinking. To refresh, the McCain camp had predicted 3 to 5 percent, I had figured the electorate would be “4 to 7 percent more Democrats than Republicans. Michael Barone said he could not rule out 8 or 9 percent. One of my most pessimistic buddies was putting it at 15 percent.”

Gallup put the Democrats’ party ID advantage at 12 percent. Rasmussen had it at 6.5 percent, and nailed the percentages for Obama and McCain.

Terrye on April 7, 2009 at 3:02 PM

58%. 64%. 66%. Who cares? The fact that over half of those polled like the guy is depressing enough. I’d feel a whole lot better for the future of this county if this were not the case.

Rod on April 7, 2009 at 3:03 PM

If it ain’t Rasmussen, it can be used but why? If it is Rasmussen, it is sp*t on meant for all’s use.

Sapwolf on April 7, 2009 at 2:57 PM
Corrected.

sethstorm on April 7, 2009 at 3:01 PM

Sapwolf on April 7, 2009 at 3:03 PM

sethstorm:

You know the constant references to Fox as if it were the boogey man or something are just absurd. There is msnbc, cbs, nbc, cnn, etc and on and on. Somehow Fox is considered just horrid, in spite of the fact that is has higher ratings and a larger variety of programming. What do you want? Pravda? Do you have to have complete control of the media?

Terrye on April 7, 2009 at 3:05 PM

You wrote this before I explained what facts Ed bases his claim of “cooking the numbers” so I just want to reiterate, look at the update and notice the spread of the last election was 7 points, not 16. Polls indicate Dem ID going down not up. There is no logical argument that a 16 point spread represents any statistically valid finding. The study is completely worthless.

msmveritas on April 7, 2009 at 2:56 PM

Let me add to that, if I may.
According to the NYT’s own numbers,they discarded 42 republican and independent responses and manufactured 42 additional democrat responses to come up with the percentages they wanted.

oldernwiser on April 7, 2009 at 3:05 PM

Obama enjoys 66% approval rating among those that voted for Obama with the remaining 34% going to ‘Look at the pretty colors’.

BigWyo on April 7, 2009 at 3:05 PM

Rod:

Give it time. People do not want to admit they were so completely duped. Hot Rod’s trial is coming up. Lots of stuff might come out of that.

Terrye on April 7, 2009 at 3:06 PM

Let me take a whack at this:

Why is the poll garbage?

Because of pool of respondents.

Who, where and what you ask for your poll will directly influence the result.

Example:
When I was a kid there was a Burger King commercial that cited something like 3 out of 4 people preferred the Whopper to the Big Mac.

However, in the fine print it stated that the only people asked where Burger King customers.

By polling only at BK, they skewed their numbers to arrive at the conclusion they wanted – that more people like their product than McDonald’s.

A very similar thing happened here.

The NYT purposefully chose more (a higher percentage) of Democrats and Independents, thus skewing the results higher for Obama.

Now, while we can’t know what was in their hearts when they did this…

…there had to be discussion about using the higher percentage and they MUST have known that by doing that they would skew the favorable rating higher.

Religious_Zealot on April 7, 2009 at 3:08 PM

Not space aliens, Lo…Chupacabras.

Mike D. on April 7, 2009 at 12:30 PM

Cut out that talk of illegals.

You are gonna spook my production line.

Sapwolf on April 7, 2009 at 3:01 PM

Coming up for vote by the legislature here in MA is giving the right to vote in state elections to illegal aliens. Really.

Rod on April 7, 2009 at 3:09 PM

The fact that over half of those polled like the guy is depressing enough. I’d feel a whole lot better for the future of this county if this were not the case.

Don’t give in to despair. Asking the opinions of insane people will yield insane results. As expected. The reality everyone is overlooking is that a large percentage of voters did not go and vote at all.

dogsoldier on April 7, 2009 at 3:12 PM

Terrye on April 7, 2009 at 3:05 PM

Fox has variety only if you want right wing.

sethstorm on April 7, 2009 at 3:13 PM

oldernwiser on April 7, 2009 at 3:05 PM

Thank you, that is an excellent point. This is even greater direct evidence of cooking the numbers. I have to say those are pretty desperate tactics too. They must have really not liked the way results were actually trending.

Here is the really sad part, MSM knows how to pick apart validity in a study. Watch them do it when the numbers don’t favor the position they want to assert. Chris Matthews voted for Bush in 2000 and I watched him pick apart these polls before. That these “reporters” mention none of this and report the findings of this study with a straight face is just absurd.

msmveritas on April 7, 2009 at 3:16 PM

sethstorm:

BS. No doubt Fox is willing to give time to people like Beck, but he has an audience, no one has to listen and look at his ratings compared to Olbermann. I could say they have a variety of programming if you are a dimwitted leftie too.

George what’s his face who used to be part of Clinton’s administration is on ABC. Matthews used to be a Democratic operative and yet I am supposed to believe he is a journalist. The late Tim Russert worked for Cuomo..the list goes on and on.

Combs is getting a show on Fox. Shep Smith is not a conservative and besides, like I said do you want to control the entire media?

Terrye on April 7, 2009 at 3:17 PM

sethstorm thinks there are two kinds of journalists, the kind who work for or used to work for the DNC and wingnuts whose opinions are not valid because their politics are not to his liking.

Terrye on April 7, 2009 at 3:18 PM

AMAZING that these dunces in the MSM don’t realize that the country is middle-right. If they just told the truth, they’d dominate network news, just as FOX has dominated cable news. So dumb.

marklmail on April 7, 2009 at 3:21 PM

According to this poll, Michelle Obama only has a 50% favorable rating. I thought she was super popular.

Rose on April 7, 2009 at 3:22 PM

Foxnews ratings and the deaths of Air America, NY times and CNN are the best polls one can look at. And the steadily declining ratings of the alphabet networks.

People KNOW socialist propaganda when they see it.

dogsoldier on April 7, 2009 at 3:29 PM

Traffic Cop Timmy on April 7, 2009 at 1:00 PM

Stupid doesn’t know it’s stupid…

ladyingray on April 7, 2009 at 3:30 PM

This is the breakout in their demographics on page 23:
Democrats – 39%
Republicans – 23%
Independents – 30%

I question even these numbers…………..

Seven Percent Solution on April 7, 2009 at 3:31 PM

msmveritas on April 7, 2009 at 3:16 PM

You’re welcome

Do not be surprised if their next poll eliminates the back up data.

oldernwiser on April 7, 2009 at 3:36 PM

Terrye on April 7, 2009 at 3:17 PM

Controlling the media? I’m not from Fox.

At least I’m not supporting an organization whom has no problem with letting a host spam a competitor.

sethstorm on April 7, 2009 at 3:43 PM

Even Kos’ polls faired better than CBS/NYT during the presidential election–very telling indeed….

RepubChica on April 7, 2009 at 1:46 PM

Oh man, that is totally rubbing their noses in it.

Way to go Chica. I didn’t think you could get lower than KOS.

By the way, some of these trolls on this thread are kinda weak. But as long as they don’t post someone’s resume I could live with them a little longer.

Sapwolf on April 7, 2009 at 3:50 PM

Fox has variety only if you want right wing.

sethstorm on April 7, 2009 at 3:13 PM

Seth, they have leftists too. You are not looking hard.

And besides it’s moot.

They have Megyn Kelly and that’s good enough for me.

Sapwolf on April 7, 2009 at 3:55 PM

oldernwiser on April 7, 2009 at 3:36 PM

Any legit research would not only support the data but also point out potential flaws. This is just SOP to give your research added credibility. Eliminating back up data would be an admission the study can’t be replicated. That said, I would not at all be surprised since obviously credibility is not the foremost goal here.

msmveritas on April 7, 2009 at 4:38 PM

NYT/CBSRassmussen poll has its thumbs on the scale … again

Corrected.

Rassumssen on the other hand has been asking questions in some very slanted ways that skew GOP. Thankfully we have plenty of pollsters to choose from that are more reliable, including even the NYT/CBS poll, believe it or not.

No wonder Fox quotes Rassmussen.

sethstorm on April 7, 2009 at 1:30 PM

So, does Rasmussen have a bad track record as far as accuracy is concerned? Hmmmm?

ddrintn on April 7, 2009 at 4:38 PM

Comment pages: 1 2 3