Slate: What is a "likely caucus goer"?

[rinsd]

"Poll Dance: A Newsweek poll released over the weekend has survey wonks chattering over who's beating whom in Iowa. The big (supposed) news: It gives Hillary a six-point lead among Democratic voters, but has Obama leading by four points among "likely caucus-goers." Some bloggers have expressed doubts about the poll given its gaping margins of error (+/- 7 percent for likely Democratic caucus-goers; +/- 9 percent for their GOP counterparts). But it also raises the perennial question: How do you define a "likely caucus-goer"?

Depends who you ask—there are virtually as many definitions as there are pollsters. In the Newsweek poll, they asked each subject to rate themselves on a four-grade scale of definitely attending, probably attending, probably NOT attending, and definitely not attending. They then define a "likely caucus-goer" as someone who said they are either "definitely" or "probably" attending.

Some polls will also factor in whether a respondent attended previous caucuses. For example, a poll conducted by Time/SRBI last August defined a "likely voter" as someone who said they were either, "100% certain that they would attend the Iowa caucuses" or were "probably going to attend and reported that they had attended a previous Iowa caucus." The Newsweek poll reports that among likely Democratic caucus-goers, 64 percent attended a previous caucus, but it doesn't factor this into its "likely caucus-goer" definition.

So why can't they just standardize the definition? Because no one's sure which methodology best reflects reality. Iowa polling is already a crapshoot since the caucus process is so complicated. For example, there's no way to simulate the rule that caucus-goers whose candidates get less than 15 percent support in a given precinct have to throw their weight behind a more popular candidate. Results also vary depending on whom pollsters talk to. Newsweek used a random digit dial sample; other surveys draw from lists of registered voters. Plus, turnout depends largely on each candidate's ground organization—if Obama gets students to turn out in record numbers in Iowa, that could throw off polling accuracy. Of course, you can't even begin to gauge organization levels until closer to the primaries, and even then it's hard to measure."

More at the link - http://www.slate.com/id/2175053/

Poll junkies definitely click the link which will take you to Slate which has links to pollster.com's Disclosure Project in which they are attmepting to get polling firms to share their methodology.

[wyldwolf]

1. in other words, the Newsweek poll was shoddy.

[rinsd]

4. 7pt MOE is not good.

[wyldwolf]

8. and letting people define themselves for the criteria the pollster is seeking

[Debi]

2. You could ask us Iowans....

:shrug:

[movonne]

3. I'm asking..how do you see it...

[Debi]

15. right now 1/3 of the 'caucus goers' that are responding to the polls

are either young first time caucus attendees or new to the party caucus attendees. There is no guarantee that they will make it to the caucus or that they will remain loyal to whichever candidate got them involved in the first place. So you have to look at the poll participants and think '1/3 of these results are fluid'. Which is why the last three polls have come out so different and each poll after that will be different too. :shrug:

(as for independents or republicans switching over to caucus. I don't think it happens as often as is being talked about. In my precinct only two republicans switched over to caucus in 2004 and none in 2000. I am from a small precinct, so maybe there was a higher turnover elsewhere).

Probably didn't help you much. But if you want to know how Iowa DUers are seeing the race head over to the DU forum. We've had several polls about it.

:hi:

[movonne]

17. thank you..

[Ethelk2044]

5. We know all the polls with Hillary leading are shoddy. We have known that for a while.

[rinsd]

10. LOL. That this is posted with a straight face makes this all the funnier.

[DemocratSinceBirth]

18. It's Good To See The Luddites Are Represented Here

om

[jenmito]

6. Why the doubt of the poll that shows Obama ahead...

yet Rasmussen, a RW telephone pollster is constantly being posted here to show Hillary running away with the nomination? Perhaps when Dems. were asked if we're ready for a Mormon president they ssid no because they were thinking specifically of Romney!

[rinsd]

9. Perhaps because that poll has an MOE of 7pts.

Rasmussen is posted here most often because he is the most prolific pollster. He has a daily tracking polls for national heats; every two weeks he does one of the big 3 primary states; he does head to head twice a week with different candidates.

Right there is at least 3 to 4 polls a week that will be of interest to DUers.

[jenmito]

11. They just showed a Newsweek poll with a MOE of 3 pts.

It looked like the same poll, Obama 28% to Clinton's 24% and Edwards' 22%. :shrug:

And Rasmussen IS a regular on Faux. I think he has an agenda.

[rinsd]

12. The overall MOE is 3 pts, Dem voters - 5 pts and Dem likely caucus goers - 7pts.

http://www.msnbc.msn.com/id/21036143/site/newsweek/%20=%20POLL%20NUMBERS

The NEWSWEEK Poll was conducted by Princeton Survey Research Associates International on Sept. 26-27. Telephone interviews were conducted with 1,215 Iowa registered voters; the overall margin of error is plus or minus 3 percentage points. The margin of error for questions asked only of Democratic voters is plus or minus 5 percentage points and "likely" Democratic voters is plus or minus 7; for Republican voters it is 6 percentage points and plus or minus 9 points for "likely" GOP voters. For respondents who said the issue of abortion is important, the margin of error is plus or minus 4 points; for respondents who said it was "not important" it's 5 points.

[zulchzulu]

7. In a nutshell, a "likely caucus goer" is someone passionate enough about...

..their preferred candidate that they would take the time and inclination to go attend a caucus, which in Iowa, would be in the middle of winter at a location for a few hours of contention.

A way to gauge a "likely caucus goer" for polls would to not only ask who they support but on a level of 1 to 10. Of course, you would ask if the person would caucus for a candidate as well.

I'll be Iowa as a caucus observer...can't wait!

[Debi]

16. Do you know where you'll be?

Have you seen a caucus before (other than the craziness shown on C-SPAN)?

How exciting!!

[zulchzulu]

19. I was in Des Moines in 2004

I'm probably going to be there from the weekend before the caucus (whenever that is now :hi:) and stay until the day after. I hope to travel around the state the week before following the campaigns and filming what I can.

I love how the caucus works.

[Debi]

20. That's awesome! I do too

Coming from Montana into Iowa politics it's quite a change! :bounce:

[killbotfactory]

13. They asked them if they were going to caucus

Then they asked those where said they were likely to go caucus who their top two choices were.

That's how they added up the "likely caucus goer".

[rinsd]

14. Different firms have different criteria.

Some ask only if you are planning to caucus, some ask in addition to that if you have caucus'ed previously. The thought being that a previous caucus goer is more likely to go than a 1st timer.

[CK_John]

21. IMO. they need to do a personality profile to get an accurate pic of a caucus goer.

Like are you passive aggressive type, are you a go along to get along type, are you a never change my opinion type, are you an avoid confrontation type, are you convinced easily by others type, are you convinced by the last person type. This info would be a better judge of the outcome.