Latest Rasmussen Report 05/05/07

[Alamom]

http://rasmussenreports.com/public_content/politics/top_stories__1/president_bush_job_approval


President Bush Job Approval
Saturday, May 05, 2007

Forty percent (40%) of American adults Approve of the way that George W. Bush is performing his duties as President.
Fifty-eight percent (58%) Disapprove of his performance.
These figures include 17% who Strongly Approve and 43% who Strongly Disapprove .
The President earns approval from 78% of Republicans but only 13% of Democrats.
Among those unaffiliated with the two major parties, 34% approve.


Can this be right? Odd, since a Newsweek Report dated 05/05 has his approval @ 28%...
(which I like much better, btw)

Can we trust any of these reports or polling numbers? Which one?



http://sev.prnewswire.com/magazines/20070505/NYSA00305052007-1.html

[Catch22Dem]

1. If you take the average of 78 + 13 + 34 you get 41%

Surely that's not the right way to arrive at his overall approval numbers is it? Pollsters and statisticians, can you help me here?

[Selatius]

2. Rasmussen is the outlier.

The best thing to do is to drop the high and low number and average all the other numbers left together. To be sure, he's somewhere in the 30s, perhaps low 30s. Incidentally, the Nazis won around 33 percent of the popular vote in the 1933 German elections.

[hobbit709]

3. The three ways of lying

Lies, damned lies and statistics. Rasmussen polls are biased in favor of numbnuts. It's all in how the questions are asked and in the range of answer choices you are given.

[Fovea]

4. Pravda

Rasmussen has become a house organ of the Neocons, IMO.

[Selatius]

5. Heh, I thought FOX News cornered that market, but I guess there's room aplenty. n/t

[achtung_circus]

6. Pollkats is a good site,

check out the approval results. His use of scatter graphs is instructive. Rasmussen won't let him include their data in his analysis.

http://www.pollkatz.homestead.com/

Why? Hmmm.

And the Flush Bush index is a hoot.

[wyldwolf]

7. Some on DU swear Rasmussen is the most accurate... so...

... live by Rasmussen and die by it.

[hlthe2b]

8. Really? Rasmussen and Gallop have always been blatently

biased..

I don't recall a major push on DU to rely on Rasmussen...:shrug:

[wyldwolf]

9. everytime a new Rasmussen comes out, someone swears it is most accurate

[hlthe2b]

10. newcomers...

perhaps? :shrug:

Rassmussen is notoriously biased towards REPUGS in both methods and "interpretation"

[wyldwolf]

13. oh, no. One LONG TIME member in particular

[tritsofme]

19. Their past performance indicates that they are biased

in favor of the actual results of elections.

For whatever that's worth.

[johnaries]

14. Actually, if you subtract 5 - 10 points from Repukes and add it to Dems

they're usually pretty close.
:rofl:
I remember when they admitted they changed their "weighting" towards Republicans because there are more Republicans than Democrats.

They also do a 3-day average. So, if ratings change overnight because of some new event they are still averaging in the old ratings.

[skipos]

11. Go look at the 2006 elections. Who was more accurate than Rasmussen?

Rasmussen's approval/disapproval are more nuanced than other polls. There are 4 choices, not 2. It might have something to do with their outlier status on Bush ratings. Or maybe their methodology sucks.

[johnaries]

15. They were also the most accurate in the 2004 election,

which we KNOW was fixed.

What does that tell you?

[skipos]

16. It tells me that we should pay attention to their polls. nt

[johnaries]

30. Because their results matched other results KNOWN to be inaccurate?

Oh, yes, I think we probably should watch their polls.

[Colobo]

17. "which we KNOW was fixed"

Nah... Kerry lost.

[johnaries]

31. Whether Kerry lost or not, the fact remains

that the Rasmussen Polls closely matched "official results" that were later found to be flawed because of glitches such as vote switching, more votes being counted for a district that people living in that district, etc, etc. The evidence showing that the official vote was flawed and inaccurate is over-whelming. And yet, Rasmussen closely matched that flawed result.

An imortant thing to remember is that Rasmussen did not EXACLTY match (which would cause me to don my tinfoil hat), but that Rasmussen was the closest.

Since we know that the "official, certified" results were skewed and inaccurate toward Republicans, the only assumption that we can logically make is that Rasmussen was also skewed towards Republicans.

If you like, I can also make a lot of other comments about their methods. I have issues about their "weighting", i also have issues about their using a 3-day average and yet reporting the results daily (this is a cheap way of inflating their sample size by 3, but in reality it increases the margin of error by 3. Not professinal, at ALL! by any statistician's standards).

Rasmussen, although popular, is not accurate and should be IGNORED!

THEY SUCK!!!!!

[tritsofme]

20. I guess it means Scott Rasmussen has a special red phone

to Karl Rove so he can trick everybody into thinking he has accurate polls?

[johnaries]

32. Well, remember when all the polls said Bush was losing....

ok, I forget - was it Rove or Cheney who said "the information I have says you're winning".

Obviously, whoever it was had inside information that the rest of the country didn't have or wasn't paying attention to.

Do you really think he was talking about the Rasmussen polls as opposed to every other scientific poll in the country?

If so, I have a bridge to sell you....

[tritsofme]

33. That's just not true.

An aggregate of polls on election night gave Bush a statistically insignicant lead.

[rinsd]

24. Rasmussen has very accurate election polls, approval numbers not so much

3 elections in a row both he and Survey USA came the closest to the final tally.

But on approval numbers I believe that he offers more choices than simply approve or disapprove.

[McCamy Taylor]

12. Fox has him at 38, and we know that they are biased.

He is still a failed president.

[alfredo]

18. From Poll Katz

This will give you an idea of the spread in polls, and if you spend some time you can see who biased for or against bush, and who seems to go the middle path. Even then there is some problems. It is an art, not a science. From this graph, one could say he has leveled out in the 30's since the end of the year.


http://www.pollkatz.homestead.com/

[Roland99]

21. $20 if you overlaid a graph of gas price increases, you'd find quite a correlation

:)

[alfredo]

22. Someone did that back when we had the Katrina price surge.

Here's one for you.



Fox News would say, "If you supported bush, gas prices would go down."

donate the $20 to DU. :)

[Roland99]

23. Consider who is likely answering the phone for a surveyor...

Someone willing to answer a landline and remain on the phone.

Seriously doubt it would be a member of the elite or Bush "Rangers" or the like.

[alfredo]

25. I dropped my land line but I can still get incoming calls because I am

using DSL. I can call out for 10 cents a minute.

[Roland99]

26. My landline is Vonage. Not sure how that fits into marketing/polling calls but

I never had a problem with telemarketers with that exchange/number.

[alfredo]

27. the no call list works well for me.

[Roland99]

28. Yeah, I added mine and my cell phone to be safe. Plus,

all of the numbers of my family :-)

[alfredo]

29. the only spam calls are from politicians.