Why Do DU'ers Continue To Cite Rasmussen Polls...

[DemocratSinceBirth]

The man is a discredited clown.... He had Bush beating Gore by seven percentage points on Election Day 2000...

He had to close down his Portrait Of America poll because it became such a farce....

[SaveElmer]

1. He's like a bottom line indicator

Since he always leans to the Republican side of things...it;s like getting the worst case scenario. If he has Kerry beating Bush, you know Bush s in trouble.

[mobuto]

2. Because a poll's a poll

A rolling auto sample should work, theoretically, and his results for 2002 were accurate. Is his polling accurate in 2004? Maybe yes, maybe (more likely) not. But I don't think you can automatically dismiss it. I think its probably most useful when you're trying to capture trends - even if the absolute numbers are befuddled.

[DemocratSinceBirth]

3. His Polling Consits Of An Automated Machine

asking you to vote 1 for Bush and 2 for Kerry...

It does nothing to qualify who is answering the phone...

It could be eight year olds answering...


The man is a discredited clown...

[mobuto]

4. Not quite

Rasmussen does use auto callers, but it isn't a simple Bush/Kerry option. Responders have to answer 30 questions, and the response is whittled down to a group Rasmussen considers likely voters. Could an 8 year old manipulate the response? Sure. But all polls are vulnerable to dishonest answers.

You don't think an unregistered voter ever tells Gallup that they're actually registered and sure to vote?

http://www.rasmussenreports.com/methodology.htm

[k in IA]

5. I got one of these calls last week. I didn't know who it was from though.

I used to get calls from Zogby when I lived in KS. The worst is when they give you crappy options and none of the responses correspond with what probably 25-50% of people would say.

[DemocratSinceBirth]

7. Zogby Uses Real People....

NT

[k in IA]

9. yes, and they will actually talk to you (especially if they are

bored). I always would try to get some info out of them and tell them when the options sucked.

Since I am a news junkie, I am always delighted when someone actually asks for my opinion.

[DemocratSinceBirth]

6. What Kind Of Person Discusses Their Vote With A Disembodied Voice?

And why did he have Shrub beating Gore by seven percentage points on Election Day??????


Another clown is Ed Goas (sp) His Battleground Poll had Bush beating Gore by five percentage points on Election Day.... In fact his poll never had Gore ahead....

By the way I did post grad work in Poli Sci and took several methodology courses....

[k in IA]

8. Well, I agree, it would be silly to "discuss" anything with a computer

program where the appropriate response is "press 1" or "press 2".

[mobuto]

10. Well, I would.

All telephone polling has the same problem - the kinds of people who would discuss politics with a complete stranger over the telephone or the kind who would punch their presidential preference into a keypad - are not representative of the public.

You might find that elderly women, for example, are grossly overrepresented. Polls therefore have to try to control for demographic groups and the results are extremely imperfect.

And why did he have Shrub beating Gore by seven percentage points on Election Day??????

I don't know. But I can guess. My guess is that a combination of two problems. Since 2000 was Rasmussen's first poll, he either screwed up his polling methodology, the definition of a "likely voter" or both. Even well established polling organizations often differ strongly on what they consider the likely voter pool, and new pollsters are often way-off.

By the way I did post grad work in Poli Sci and took several methodology courses....

Well then what am I doing lecturing you? You should be lecturing me. Get to it. ;-)

[DemocratSinceBirth]

11. Gallup, Newsweek, Zogby,Harris, CBS News Were All Within

the margin of error in 00...

Their polls tend to be the most reliable though Zogby and Gallup are erratic at time....

[k in IA]

13. Zogby was the only one that got the 2000 election correct

[DemocratSinceBirth]

15. CBS , Harris, and Newsweek Nailed It Too....

As I remember Zogby declared Gore the winner by 2% in 00....


Zogby caught the late trend for sure but Gore won by .75%

[k in IA]

16. After the FL debacle, we know that what people said they did or will

do (in a poll) is not always what the ballots show.

The FL exit polls probably did show Gore winning but the butterfly ballot in West Palm Beach with Buchanan getting that huge Jewish vote was probably just a bit different from what people leaving the polls said they did. Aarrgghhh!

[MoonRiver]

17. What happened to Harris?

I haven't seen it for a long time.

[k in IA]

12. Both Zogby and Rasmussen asked age, gender, party ID, etc

questions.

By the way, I'm not elderly - I'm 34 with three small children who amazingly must not have been running around screaming at the time of the call.

[DemocratSinceBirth]

14. You Have To Control For Race, Age, Party, Gender,Etcetera

or else your polls are meaningless....


In other words you need to get a representative sample that represents the population you are trying to poll....


In other words you need the same percentage of African Americans, Caucasians, Dems, Reps, etcetera...


That's where I believe Rasmussen went astray in 00

Obviously, Reps and Caucasians were overrepresented in his polls...