Reported ARG poll numbers in ERROR, spread the word!

[JJ2]

The overall numbers for the ARG poll, as reported on their website, are in error!

If you look at the detailed numbers (here: http://americanresearchgroup.com/pres2012/primary/rep/nh/), it says that "Definite" voters make up 97% of their "likely voters" polled and "Probably" make up 3%. And yet Paul is beating Huntsman 23%-11% among "Definite" voters and Huntsman is beating Paul 24%-8% among "Probably."

So someone explain to me how they get 17% to 16% overall? It looks like their overall numbers are a 50/50 split of the "Definite"/"Probably" numbers, instead of a 97/3 split.

Comparing it to their SC poll (http://americanresearchgroup.com/pres2012/primary/rep/sc/), it looks like they really made a mistake here. In SC, the "Definite/Probably" breakdown is 90%-10%, which is much more in line with a 97%-3% split than a 50/50 split.

I don't think they would do something so obvious on purpose. (But who knows?!)

If these numbers really are erroneous, and it looks that way to me (especially comparing it with other recent polls), then we need to get the word out FAST!

[redmod79]

That's definitely what they did. P.O.S. pollsters.

[RickyJ]

The overall numbers for the ARG poll, as reported on their website, are in error!

If you look at the detailed numbers (here: http://americanresearchgroup.com/pres2012/primary/rep/nh/), it says that "Definite" voters make up 97% of their "likely voters" polled and "Probably" make up 3%. And yet Paul is beating Huntsman 23%-11% among "Definite" voters and Huntsman is beating Paul 24%-8% among "Probably."

So someone explain to me how they get 17% to 16% overall? It looks like their overall numbers are a 50/50 split of the "Definite"/"Probably" numbers, instead of a 97/3 split.

Comparing it to their SC poll (http://americanresearchgroup.com/pres2012/primary/rep/sc/), it looks like they really made a mistake here. I don't think they would do something so obvious on purpose. (But who knows?!)

If these numbers really are erroneous, and it looks that way to me (especially comparing it with other recent polls), then we need to get the word out FAST!


The media does not care about the truth here, neither does the polling company. Telling the media about this will do no good. I highly doubt that any poll we are getting these days from the establishment is accurate, they just can't be trusted. If we want to know the truth we are going to have to conduct our own poll.

[Thomasjefferson69]

Find contact info for arg and we can mail them.

[sailingaway]

Can someone who is better at math than I am (randomly, any other person on these forums) double check that? Because Huntsman has already been tweeting that stupid poll out...

[JJ2]

They don't have an email address listed only a phone number: 603/624-4081

I hope someone can alert the Paul campaign to this. They could at least put out a statement on this or make ARG correct it

[QWDC]

Didn't they do something like this the last poll, where Paul polled 25% with men and 17% with women and he got 17% overall?

[redmod79]

RP: 23*.97 + 8*.03 = 22.55

JH: 11*.97 + 24*.03 = 11.39

[sailingaway]

RP: 23*.97 + 8*.03 = 22.55

JH: 11*.97 + 24*.03 = 11.39

thank you!

[lucent]

Someone that knows the calculations, call them.

603/624-4081 :)

[Lord Xar]

Who owns a r g?

[69360]

What in the world do these numbers mean? The 10 and the 7-9? On a scale of 1-10???


Likely Republican
Primary Voters Definite - 10 (97%) Probably 7-9 (3%)

[Liberty74]

By looking at the breakdown of Republican voters, Independent and Tea Party for each candidate I don't see how Huntsman is in second. The only possible way is if they polled more Independents than Republicans, right?

Registered Republicans:
Romney 46%
Paul 14%
Huntsman 12%
Santorum 10%

Independents
Romney 30%
Huntsman 24%
Paul 20%
Santorum 15%.

Tea Party Supporters
Romney 33%
Gingrich 22%
Paul 15%
Santorum 15%

So to get a Santorum surge, CNN and the media whores that ran with it left out Independents. To get a Huntsman surge, ARG included a very high amount of Independents?

[Hoax]

They don't differentiate between definite and probably. So they did:

RP: 14*.6 + 20*.4 = 16.4 = 16

JH: 12*.6 + 24*.4 = 16.8 = 17

Nothing scandalous here. Aside from the BS poll.

[RDM]

It would not surprise me if there are some "anonymous" donations being made to this company in return for some "accidental" mistakes.

[lucent]

They don't differentiate between definite and probably. So they did:

RP: 14*.6 + 20*.4 = 16.4 = 16

JH: 12*.6 + 24*.4 = 16.8 = 17

Nothing scandalous here. Aside from the BS poll.

They did differentiate in the SC poll.

[Hoax]

They did differentiate in the SC poll.

Well in that case, game on.

Edit: No, I don't think they did. Look at the SC overall numbers, its based on Rep * % + Ind * %. No split between definite and probable.

[JJ2]

What in the world do these numbers mean? The 10 and the 7-9? On a scale of 1-10???

Yes, at the bottom of the page it says that it's a scale of 1-10. I'm guessing that only 7-10 answers make the cut overall for their poll of "Likely Republican Primary Voters."

[JJ2]

Well in that case, game on.

Yes, the Independent vs. Republican, Tea Party, etc. breakdowns look like they're based on the false 50/50 split of definite/probably voters.

[redmod79]

If you look at the other tabs, the results would be different though.

Perferences by Party:
RP: 14*.60 + 20*.40 = 16.4

by Tea Party support:
RP: 15*.27 + 17*.73 = 16.46



So something is fishy. The results for all candidates is about the average of the likely to vote splits. Did they have a typo for the likely to vote %? I don't know.

[JJ2]

If you look at the other tabs, the results would be different though.

Perferences by Party:
RP: 14*.60 + 20*.40 = 16.4

by Tea Party support:
RP: 15*.27 + 17*.73 = 16.46



So something is fishy. The results for all candidates is about the average of the likely to vote splits. Did they have a typo for the likely to vote %? I don't know.

No, that's the thing. In their South Carolina poll, the split is 90-10 definite/probably voters, and the numbers add up correctly: http://americanresearchgroup.com/pres2012/primary/rep/sc/

[Diurdi]

I can't access either of their pages.

Forbidden

You don't have permission to access /pres2012/primary/rep/sc/ on this server.

Additionally, a 403 Forbidden error was encountered while trying to use an ErrorDocument to handle the request.

[lucent]

JJ2, call them. You know what you are talking about.

[JJ2]

JJ2, call them. You know what you are talking about.

I'm not able to call right now, but I'll see what I can do.

[redmod79]

Either the likely to vote split is wrong or all of the other splits are. Perhaps they fudged the numbers and forgot to fudge the likely voters split.

[floydmoon]

Website is not accesible to me.

[rp2012win]

I called and left a message.

[JJ2]

I found an email address for them through WHOIS: arginc@AOL.COM

[lucent]

I found an email address for them through WHOIS: arginc@AOL.COM

Call and email. Doing both will have a bigger impact.

[maxoutco]

Here's the deal. Someone from the campaign must interview w/ the MSM and state these blatant lies and say will you do our nation a favor and investigate how these polls are working and show to them that it is wrong on live t.v.

[The Gold Standard]

Of definite voters the poll is:

Romney 35%
Paul 23%
Gingrich 12%
Santorum 11%
Huntsman 11%
Undecided 6%
Other 2%
Perry 0%


Of probable voters the poll is:

Romney 45%
Huntsman 24%
Santorum 13%
Paul 8%
Undecided 3%
Gingrich 3%
Perry 3%
Other 1%


If defniite voters really made up 97% of this poll, the real results are:

Romney 35.3%
Paul 22.6%
Gingrich 11.7%
Huntsman 11.4%
Santorum 11.1%
Undecided 5.9%
Other 2.0%
Perry 0.0%

[Liberty74]

Since the Paul campaign cannot keep up with the media BS, Lies and Smear to refute with a press release (free news coverage), does the campaign need to have a 800 hotline where we the grassroots can keep them informed so they can be on top of it all? Just a thought...this ARG poll is an example. And maybe if the campaign got out in front of that CNN poll, we would be sitting differently?

I think ARG got their definite and probably numbers mixed up. Unless we know the break down of the Republican v Independent vote in their poll (80/20, 70/30, 50/50) we can't say either way. Does anyone know those numbers?

[WD-NY]

RP: 23*.97 + 8*.03 = 22.55

JH: 11*.97 + 24*.03 = 11.39

This needs to be tweeted to all the major reporters. (speaking of which, we should have crib sheet that lists the twitter & email addresses of all the major reporters pinned at the top of grassroots central)

[Jack Bauer]

Don't jump the gun.

Their numbers only make sense if the Likely/Probably Split is 55-45 and NOT 97-3 as reported.

Other numbers (Repub/Indy split and Tea Party/Non Tea Party split) are consistent.

http://img35.imageshack.us/img35/2792/uploadz.png

[RickyJ]

Since the Paul campaign cannot keep up with the media BS, Lies and Smear to refute with a press release (free news coverage), does the campaign need to have a 800 hotline where we the grassroots can keep them informed so they can be on top of it all? Just a thought...this ARG poll is an example. And maybe if the campaign got out in front of that CNN poll, we would be sitting differently?

I think ARG got their definite and probably numbers mixed up. Unless we know the break down of the Republican v Independent vote in their poll (80/20, 70/30, 50/50) we can't say either way. Does anyone know those numbers?

There is nothing we or Ron can do to combat these polls unless we do our own polls, get vastly different results, and then tell everyone that something is wrong with all these polls. Let's fact it, we have so many powerful organizations trying to stop Ron Paul from becoming president that it will take a miracle for him to become president.

[JJ2]

I think ARG got their definite and probably numbers mixed up.


Don't jump the gun.

Their numbers only make sense if the Likely/Probably Split is 55-45 and NOT 97-3 as reported.

Other numbers (Repub/Indy split and Tea Party/Non Tea Party split) are consistent.

Except their South Carolina poll would indicate the definite/probably breakdown is accurate. It's 90%-10% split in SC: http://americanresearchgroup.com/pres2012/primary/rep/sc/

[RonPaulRules]

http://www.timothyjwelsh.com/courses/11fa394/files/2011/09/i-see-what-you-did-there.jpg

[JJ2]

If defniite voters really made up 97% of this poll, the real results are:

Romney 35.3%
Paul 22.6%
Gingrich 11.7%
Huntsman 11.4%
Santorum 11.1%
Undecided 5.9%
Other 2.0%
Perry 0.0%

Yes, which just happens to be totally in line with other recent polls!

[Jack Bauer]

Except their South Carolina poll would indicate the definite/probably breakdown is accurate. It's 90%-10% split in SC: http://americanresearchgroup.com/pres2012/primary/rep/sc/

I don't know what the Likely-Probably split is. Only ARG knows that.

If it is 97-3 as they self reported, then their numbers are grossly screwed up.

However, if they didn't screw up, the Likely-Probably split should be closer to 55-45.

[Diurdi]

I don't know what the Likely-Probably split is. Only ARG knows that.

If it is 97-3 as they self reported, then their numbers are grossly screwed up.

However, if they didn't screw up, the Likely-Probably split should be closer to 55-45. So either their Likely-Probably split number is false, or their poll numbers are false.

Either way, what a massive fuck up of a pollster.

[Dianne]

Said it a million times on this forum... Time for attack ads on the news media... expose their lying, asses... Forget about the other candidates... expose the news media for making GODS out of crooks.. make people awaken to the brainwashing, and they will be free to vote for Ron Paul.

The truth shall set them free.

[Jack Bauer]

So either their Likely-Probably split number is false, or their poll numbers are false.

Either way, what a massive fuck up of a pollster.

Yup. Pretty much.

[RickyJ]

So either their Likely-Probably split number is false, or their poll numbers are false.

Either way, what a massive fuck up of a pollster.

I think the whole poll is false.

[Johnnybags]

LOL, Boston media reporting Huntsman surge into second, media is so corrupt, we live in Russia.

[JJ2]

I think think whole poll is false.

And the worst part is that Huntsman is about to go on national TV for the debate and tell everyone he is in 2nd place now!

[Liberty74]

And the worst part is that Huntsman is about to go on national TV for the debate and tell everyone he is in 2nd place now!

And that is the whole purpose. To artificially create a surging candidate to knock Paul out of the top 2 spots. It worked in Iowa. The powers to be know what they are doing with pollsters and the media.

[kill the banks]

can someone let campaign know please

[Jack Bauer]

Someone that knows the calculations, call them.

603/624-4081 :)

Goes to their voice mail.

[Tyler_Durden]

Yup. Pretty much.

Bingo!! This--->
http://www.ronpaulforums.com/showthread.php?347737-War-on-Two-Fronts-The-GOP-and-the-MSM

[Tyler_Durden]

Said it a million times on this forum... Time for attack ads on the news media... expose their lying, asses... Forget about the other candidates... expose the news media for making GODS out of crooks.. make people awaken to the brainwashing, and they will be free to vote for Ron Paul.

The truth shall set them free.

Bingo! This --->http://www.ronpaulforums.com/showthread.php?347737-War-on-Two-Fronts-The-GOP-and-the-MSM

[JJ2]

Another ARG poll (for Florida) which shows a "Definite"/"Probably" breakdown of 90/10: http://americanresearchgroup.com/pres2012/primary/rep/fl/ It looks like their polls never have anywhere close to a 50/50 breakdown (as this NH poll would have to have in order for Huntsman to actually be in 2nd)

[ord33]

JJ2: Good point about Florida.

At this point I think its fairly safe to say that ARG either made an error with their 97/3 calculation or had a corrupt poll where they made stuff up out of nowhere.

[WD-NY]

lol


daveweigel (https://twitter.com/#!/daveweigel) daveweigel
"New ARG poll" deeeeEEEEEleted

[Varin]

lol

[tucker]

ARG:

"Using a scale of 1 to 10, with 1 meaning that you will definitely not vote in the 2012 Republican presidential preference primary and 10 meaning that you will definitely vote in the 2012 Republican presidential preference primary, what number would you give as your chance of voting in the 2012 Republican presidential preference primary?"

The graphic at the bottom shows only 10-7 (10 = Definite, 7-9 = Likely).

Therefore, Could the discrepancy be coming from the fact that they don't show the percentage for those who answered less than 7?

Just throwing it out there, I could be completely wrong.

EDIT: If this were the case, it wouldn't make the poll anymore valid in my opinion.

[Jack Bauer]

lol

I can still access it.

[Oddone]

This is why RCP refuses to use their polls.

[Jack Bauer]

ARG:

"Using a scale of 1 to 10, with 1 meaning that you will definitely not vote in the 2012 Republican presidential preference primary and 10 meaning that you will definitely vote in the 2012 Republican presidential preference primary, what number would you give as your chance of voting in the 2012 Republican presidential preference primary?"

The graphic at the bottom shows only 10-7 (10 = Definite, 7-9 = Likely).

Therefore, Could the discrepancy be coming from the fact that they don't show the percentage for those who answered less than 7?

Just throwing it out there, I could be completely wrong.

No, the poll is for likely republican voters i.e., those who scored from 7 through 10.

They were split into two categories - Definite voters (97%) who were rated 10 and Probable Voters (3%) who were rated 7-9 making up 100% of the sample.

[redmod79]

ARG:

"Using a scale of 1 to 10, with 1 meaning that you will definitely not vote in the 2012 Republican presidential preference primary and 10 meaning that you will definitely vote in the 2012 Republican presidential preference primary, what number would you give as your chance of voting in the 2012 Republican presidential preference primary?"


The graphic at the bottom shows only 10-7 (10 = Definite, 7-9 = Likely).

Therefore, Could the discrepancy be coming from the fact that they don't show the percentage for those who answered less than 7?

Just throwing it out there, I could be completely wrong.

EDIT: If this were the case, it wouldn't make the poll anymore valid in my opinion.

Unlikely voters wouldn't be relevant, anyways. Other polls match up with the other splits.

[tucker]

Unlikely voters wouldn't be relevant, anyways. Other polls match up with the other splits. This is exactly what I was thinking. I just wanted to cover all the bases before claiming it's flawed.

[pauladin]

Huntsman pimping the hell out of this poll on his facebook page.

[wgadget]

Huntsman pimping the hell out of this poll on his facebook page.

We need to let him in on the bad news, then.

[Liberty74]

Huntsman pimping the hell out of this poll on his facebook page.

That's because he is crapping his pants that all 50 people who show up to his little rallies on the side walk were polled.

[JJ2]

Anyone can post to his wall without even clicking "Like": http://www.facebook.com/jonhuntsmanjr?sk=wall

[ord33]

I just posted: "Too bad the ARG poll made an error in their "Definite" vs. "Probable" voter percentages that skews the entire poll. It is a farce! Take a look for yourself at the bottom of the poll (last category) and let me know how that 97/3 comes out to the percentages they are saying is the overall percentage of the candidates. http://americanresearchgroup.com/pres2012/primary/rep/nh/"

[limequat]

Tweet it to PPP, they may call them out.

[robert9712000]

I posted the error on john huntsmans facebook ,emailed the error to ppp and left a message on ARG's voice mail

[JJ2]

Tweet it to PPP, they may call them out.

Good idea.

[Avalon]

Hey, just so you guys know they removed the bottom table from the release. So I looked at the other tables and they seem to match the topline numbers. Was that true before? I didn't save a copy.

[JJ2]

Hey, just so you guys know they removed the bottom table from the release. So I looked at the other tables and they seem to match the topline numbers. Was that true before? I didn't save a copy.

Yes, the other tables look the same.

Wow, they changed the "Likely to Vote" to "Preference by Sex" (which they don't even have for the other polls, such as Iowa, SC, and FL), without any admission that they had made any error.

[The Gold Standard]

Tweet it to PPP, they may call them out.

Shouldn't have done that. That gave them the idea to do it for their poll too.

[jersdream]

HAHAHA they took off likelihood to vote. Such bullsh*t.

[PauliticsPolitics]

ARG always = Garbage

[ord33]

I have it as a .pdf if anybody wants it. Probably not that important, but if someone needs it with the original information just let me know.

[ord33]

Wow...They kept the wording the same on the questions asked part of their poll:


About this Survey -

Survey Sponsor: American Research Group, Inc.

The American Research Group has been conducting surveys of voters since 1985.

Sample Size: 600 completed telephone interviews among a random sample of likely Republican primary voters living in New Hampshire (359 Republicans and 241 undeclared voters (independents)). An "*" in the tables indicates responses of less than 1 /2 of 1 percent.

Sample Dates: January 6-7, 2012

Margin of Error: ± 4 percentage points, 95% of the time, on questions where opinion is evenly split.

Question Wording:

If the 2012 Republican presidential preference primary were being held today between (names rotated) Newt Gingrich, Jon Huntsman, Ron Paul, Rick Perry, Mitt Romney, and Rick Santorum, for whom would you vote?

Do you consider yourself to be a supporter of the Tea Party movement, or not?

Using a scale of 1 to 10, with 1 meaning that you will definitely not vote in the 2012 Republican presidential preference primary and 10 meaning that you will definitely vote in the 2012 Republican presidential preference primary, what number would you give as your chance of voting in the 2012 Republican presidential preference primary?

And don't have anything related to "sex" in their stated questions, yet it somehow replaced how likely they were to vote as one of the categories?

[JJ2]

Wow...They kept the wording the same on the questions asked part of their poll:



And don't have anything related to "sex" in their stated questions, yet it somehow replaced how likely they were to vote as one of the categories?

Wow. Great catch, thanks!