PDA

View Full Version : Why this Texas poll is great news! It demonstrates why RP polls poorly elsewhere!




Original_Intent
06-30-2007, 08:02 AM
http://ivrpolls.com/index.php?option=com_content&task=view&id=20&Itemid=1

Here is the article (minus the table) I have boldedareas for emphasis

polled past Republican primary voters in Texas on their preferences among the announced presidential candidates. They're not sure. While only 16% said they were undecided, 70% said they were still considering other candidates. When asked about Fred Thompson entering the race, enough said they would switch and support his candidacy to move him into the lead, but only a third of those that were still considering others said they would definitely support Thompson.

Thompson's highest support was male. In particular, men over 40 were almost twice as likely to support him as women over 40. Men were three times as likely as women to support Ron Paul. Other gender gaps saw McCain's female support 50% over his male support. Guiliani was slightly higher among women. Women were also 50% more likely to be undecided.


Without Thompson
Giuliani - 26%
McCain - 18%
Romney - 15%
Huckabee - 7%
Paul - 6%
Other Announced - 13%
Undecided - 15%


With Thompson
Thompson - 29%
Giuliani - 21%
McCain - 13%
Romney - 9%
Huckabee - 4%
Paul - 6%
Other Announced - 6%
Undecided - 11%

All announced candidate names were read (rotated), but only the choices shown were given as options (also rotated). I suspect that confusion between Tommy Thompson and Fred Thompson resulted in 'Other Announced' being as high as it is in the first list. Those picking 'Other Announced' were much more likely than average to switch to Fred Thompson.

736 past GOP Primary Voters polled 6/19/2007 - Margin of Error 3.6%

I should note that Ron Paul's support was relatively strong in the areas around his congressional district and very weak elsewhere. His district is close to Houston and relatively close to San Antonio and Austin, and his support was strongest in these areas.



OK, now for my points.

1. They polled past GOP primary voters - so Ron Paul's independent, third party, and Democrat support is not represented.

2. Only 30% of those polled say they are not considering other candidates

3. With Fred Thompson thrown in, every candidate except Ron Paul loses support to him. This shows that Ron Paul's 6% is rock solid (edit: sorry I just noticed on the poll that only 4% said they were DEFINITLEY supporting Ron Paul) . Since only 30% of those polled are committed to their choice, this means that 20%(edit to 12.5% due to prior error) of those that are committed are Ron Paul supporters! (and again this is only among past GOP primary voters!

4. I think it is very important to look at the methodologies of these other polls. Almost all of them are saying that they are from "likely GOP primary voters". It is critical to find out how that is determined. If they just randomly call anyone and ask if they are likely to vote in the GOP primary, that provides a good sample. However, if they are doing like this poll did, and preselecting likely GOP primary voters, then this is HUGE GOOD news to the RP cause

Wyurm
06-30-2007, 08:46 AM
Now, if only we could get a truely random scientific poll that is nationally recognized to stop with the past primary thing.

Bradley in DC
06-30-2007, 09:01 AM
The main thing all of these polls before Labor Day are going to show is a reflection of name ID. We need to get the message out!

DXDoug
06-30-2007, 09:07 AM
Nice! :p

stevedasbach
06-30-2007, 11:41 AM
Here's another way that some MSM polls are biased:

The most recent FOX News and Mason-Dixon polls started by asking respondents to rate candidates favorable or unfavorable. Only the top tier candidates/potential candidates were listed. THEN they asked who they were supporting for president from a list of ALL the candidates.

By rating all the top-tier candidates first, the respondent is effectively comparing these candidates to each other. By the time they finish, the repondent has in mind which of those candidates he/she supports. When they start reading the list of names, the respondent is predisposed to pick his/her favorite from the top-tier, unless he/she is already strongly committed to another candidate.

MozoVote
06-30-2007, 11:45 AM
It really is a disservice to keep calling some candidates "top tier". Call them what they are -- the best funded. There have been no votes yet. We do not know who is really "in front" or "on top".

Is it any mystery that the MSM fawns over the best funded candidates? Connect the dots:

AD.E.T.SI.G
R.VENU.

:)

mikelovesgod
06-30-2007, 12:44 PM
Here's another way that some MSM polls are biased:

The most recent FOX News and Mason-Dixon polls started by asking respondents to rate candidates favorable or unfavorable. Only the top tier candidates/potential candidates were listed. THEN they asked who they were supporting for president from a list of ALL the candidates.

By rating all the top-tier candidates first, the respondent is effectively comparing these candidates to each other. By the time they finish, the repondent has in mind which of those candidates he/she supports. When they start reading the list of names, the respondent is predisposed to pick his/her favorite from the top-tier, unless he/she is already strongly committed to another candidate.

Can this be verified? If so you are making a great point.

Zydeco
06-30-2007, 12:50 PM
As the pollster himself writes:

I should note that Ron Paul's support was relatively strong in the areas around his congressional district and very weak elsewhere. His district is close to Houston and relatively close to San Antonio and Austin, and his support was strongest in these areas.

Which doesn't mean people in other parts of Texas won't like RP, just that they don't know him yet.

FSP-Rebel
06-30-2007, 12:58 PM
I always thought that the polls were of past GOP primary voters and thus the skew. And once the Tancredos and Hunters are gone, we could realistically be over 10% of the GOPers--not counting the other supporters such as libertarians, constitutionalists, anti-war dems, independents and some greens. This is all very heartening I must say. Can't wait for the dollar totals to come out.

George
06-30-2007, 02:54 PM
Actually most media polls are RDD - random digit dial instead of RBS - Registration Based Sample. They call randomly generated numbers and ask screening questions, but a lot more people say they will vote than actually do.

Ron Paul probably gets screwed both ways. With RDD, for every infrequent voter that will make a point to vote for Ron, you get ten that have only heard of McCain and Giuliani, and probably won't end up voting this time either. With RBS, you get rid of the clueless, but don't add in the non-voters inspired by Paul.

kylejack
06-30-2007, 02:57 PM
Most polls poll LIKELY primary voters, not former voters. This particular poll is an aberration.

DAZ
06-30-2007, 03:06 PM
This particular poll is an aberration.

HA! This particular election is an aberration.