PDA

View Full Version : Iowa Robopoll Report 7/7/2011




rp08orbust
07-08-2011, 03:53 AM
.
Previous Iowa Robopolls

Please see the thread from 6/30/2011 for an explanation of why I'm doing these robopolls.

6/30/2011 (http://www.ronpaulforums.com/showthread.php?301333-Iowa-Robopoll-Report-6-30-2011)
7/1/2011 (http://www.ronpaulforums.com/showthread.php?301526-Iowa-Robopoll-Report-7-1-2011)
7/6/2011 (http://www.ronpaulforums.com/showthread.php?302176-Robopoll-Report-7-6-2011)


Robopolling Account

Opening Balances
PayPal: $13.19
CallFire: $9.07
Dynamic Interactive: $115.71
NeoSpeech: 422 words ($0.05 per word)

Income/Expenses
PayPal:
Sales (IDefendthePlatform)... +$200.00

Dynamic Interactive:
7/6/2011 robopolling... -$90.46

Closing Balances
PayPal: $213.19
CallFire: $9.07
Dynamic Interactive: $25.25
NeoSpeech: 422 words ($0.05 per word)


Purchase Ron Paul Phone Lists

I am currently selling phone numbers of those who identified themselves as Ron Paul supporters in robopolls; see http://www.ronpaulforums.com/showthread.php?302597-Robopoll-Project-Accounting for the current price per phone number. These phone numbers are excellent for calling to find Ron Paul supporters interested in attending the Ames straw poll on August 13.

If you want to purchase these phone numbers, you can make a payment directly to my PayPal account, r gear y iii at gmail dot com. If you describe it as a Personal Transfer from your PayPal balance or bank account, PayPal charges no fee (DO NOT describe it as payment for a good or service). Do not describe it as a purchase or else I could be charged an international transaction fee. If you cannot pay me without incurring a fee to you or myself, then consider using Amazon Payments to make the payment (my account email address is the same as my PayPal email address): https://payments.amazon.com/sdui/sdui/index.htm . In your PayPal or Amazon Payment note, please indicate your RPF username; otherwise I will use your initials in the public robopoll project accounting unless you request to be anonymous.

Each list of phone numbers you purchase is guaranteed to be unique from any other list I sell (i.e., the same phone number will not be sold more than once). Once I have conducted the robopolls necessary to fill your order, I will send the list of phone numbers as an attached Excel file to the email address of your choice (e.g., your own, Steve's, etc). You are welcome to share your lists with others.


Today's Samples

My polls today covered two different groups. The first group (A) consisted of all 5,002 voters in the 5/16/2011 Iowa voter database who met the following criteria:

1. Registered to vote in Iowa as independent (Non-Partisan).
2. Female
3. Will be younger than 40 years hold on August 13.
4. The phone number provided in the voter database is a cell phone number as determined by Interactive Dynamics.
5. Primary ZIP code is 50099 or 50073 or 50007 or 50152 or 50075 or 50161 or 50021 or 50102 or 50278 or 50156 or 50247 or 50130 or 50023 or 50226 or 50055 or 50230 or 50246 or 50212 or 50032 or 50035 or 50031 or 50132 or 50313 or 50051 or 50316 or 50317 or 50234 or 50239 or 50301 or 50302 or 50303 or 50304 or 50305 or 50306 or 50319 or 50328 or 50329 or 50330 or 50331 or 50332 or 50333 or 50334 or 50335 or 50336 or 50339 or 50340 or 50359 or 50360 or 50361 or 50362 or 50363 or 50364 or 50393 or 50394 or 50396 or 50936 or 50940 or 50947 or 50950 or 50980 or 50981 or 50347 or 50350 or 50397 or 50391 or 50367 or 50368 or 50369 or 50380 or 50381 or 50392 or 50276 or 50131 or 50109 or 50307 or 50318 or 50395 or 50310 or 50308 or 50309 or 50314 or 50271 or 50122 or 50162 or 50168 or 50107 or 50009 or 50206 or 50315 or 50320 or 50322 or 50311 or 50312 or 50324 or 50028 or 50249 or 50040 or 50158 or 50321 or 50034 or 50327 or 50120 or 50398 or 50325 or 50169 or 50323 or 50039 or 50127 or 50265 or 50111 or 50043 or 50223 or 50064 or 50005 or 50220 or 50595 or 50006 or 50259 or 50266 or 50078 or 50235 or 50258 or 50142 or 50263 or 50148 or 50129 or 50217 or 50269 or 50237 or 50047 or 50071 or 50167 or 50557 or 50054 or 50141 or 50211 or 50063 or 50530 or 50173 or 50061 or 50627 or 50544 or 50059 or 50066 or 50228 or 50118 or 50126 or 50101 or 50521 or 50038 or 50227 or 50106 or 50599 or 50621 or 50125 or 50003 or 51462 or 50532 or 50160 or 50128 or 50069 or 50543 or 50632 or 50431 or 50452 or 50609 or 50135 or 50229 or 50001 or 51459 or 50033 or 50261 or 50526 or 50050 or 50672 or 50208 or 50525 or 50277 or 52339 or 50026 or 50635 or 50225 or 50146 or 50041 or 51443 or 50421 or 50470 or 52342 or 50569 or 50523 or 50420 or 50501 or 50145 or 50166 or 50533 or 50441.
6. Were not called by rp2012.org or previous robopolls.

All of the ZIP codes are between 20 and 64 miles driving distance from the Hilton Coliseum, as determined by zip-codes.com (see http://www.ronpaulforums.com/showthread.php?300831-Driving-distances-from-Ames-by-zip-code&p=3367532&viewfull=1#post3367532). This area includes Des Moines and is just large enough to include Cedar Rapids and Fort Dodge.

The second group (B) consisted of all 8,851 voters who met the following criteria:

1. Registered to vote in Iowa as independent (Non-Partisan).
2. No gender discrimination.
3. Will be younger than 40 years hold on August 13.
4. The phone number provided in the voter database is a landline phone number as determined by Interactive Dynamics.
5. Primary ZIP code is 50010 or 50012 or 50013 or 50011 or 50201 or 50105 or 50014 or 50124 or 50134 or 50248 or 50046 or 50236 or 50056 or 50231 or 50037 or 50244 or 50243 or 50154 or 50036 or 50099 or 50073 or 50007 or 50152 or 50075 or 50161 or 50021 or 50102 or 50278 or 50156 or 50247 or 50130 or 50023 or 50226 or 50055 or 50230 or 50246 or 50212 or 50032 or 50035 or 50031 or 50132 or 50313 or 50051 or 50316 or 50317 or 50234 or 50239.
6. Were not called by rp2012.org or previous robopolls.

All of the ZIP codes are between 0 and 32 miles driving distance from the Hilton Coliseum, as determined by zip-codes.com (see http://www.ronpaulforums.com/showthread.php?300831-Driving-distances-from-Ames-by-zip-code&p=3367532&viewfull=1#post3367532). The area includes a large chunk of Des Moines.


The question

I used the same question for both groups:


Good evening. I'd like to know which candidate you currently support for president in 2012. Please enter: One for Barack Obama. Two for Mitt Romney. Three for Michele Bachmann. Four for Tim Pawlenty. Five for Sarah Palin. Six for Rick Perry. Seven for Ron Paul. Or zero for other or undecided. Once again, please enter. One for Barack Obama. Two for Mitt Romney. Three for Michele Bachmann. Four for Tim Pawlenty. Five for Sarah Palin. Six for Rick Perry. Seven for Ron Paul. Or zero for other or undecided.


Results: Group A

There were 165 responses to the poll question as follows:

Barack Obama 68 (41%)
Michele Bachmann 17 (10%)
Mitt Rmoney 12 (7%)
Ron Paul 10 (6%)
Sarah Palin 8 (5%)
Tim Pawlenty 7 (4%)
Rick Perry 1 (1%)
Other/Undecided 42 (25%)


Results: Group B

There were 267 responses to the poll question as follows:

Barack Obama 101 (38%)
Michele Bachmann 29 (11%)
Mitt Romney 24 (9%)
Sarah Palin 24 (9%)
Tim Pawlenty 13 (5%)
Ron Paul 12 (4%)
Rick Perry 3 (1%)
Other/Undecided 61 (23%)


Costs

The 13,853 dials (including up to 2 redials for unanswered and busy lines) cost $90.46 of credit at Dynamic Interactive. The 432 poll responses cost about $0.21 each, and the 22 Ron Paul supporters identified cost about $4.11 each.

The rise in cost of identifying Ron Paul supporters was predictable, though I didn't expect it to rise quite as much as it did. Sample B was the first to include landlines (and completely exclude cell phones, which have been exhausted in the area 63 miles driving distance from Ames). The use of landlines means we lose control over who we're polling, because anyone in the house can answer the phone, not just the independents under 40 Ron Paul with whom Ron Paul tends to do better. The lack of guarantee about who answers the phone is why I make no attempt at targeting men with landlines before women.


Strategy Change?

Deleted by rp08orbust.

ForLibertyFight
07-08-2011, 04:07 AM
You must spread some Reputation around before giving it to rp08orbust again.

IDefendThePlatform
07-08-2011, 06:00 AM
Awesome work man. 22 more potential straw poll/caucus votes.

rp08orbust
07-08-2011, 08:50 AM
Awesome work man. 22 more potential straw poll/caucus votes.

Thank you for your purchase!

You all ought to know that Tony is seriously going all out for Ron Paul--the above purchase is just the tip of the iceberg.

rp08orbust
07-08-2011, 10:39 AM
If you want to ensure that robopolling is funded through the weekend, now is the time to make a purchase.

trey4sports
07-08-2011, 10:59 AM
Is there anyway to know how much overlap there is between the names you are Identifying and what the campaign already has? Have you had Steve crosscheck a sample of your list vs. the campaigns own list to see what % of your list is fresh ?

rp08orbust
07-08-2011, 11:01 AM
Is there anyway to know how much overlap there is between the names you are Identifying and what the campaign already has? Have you had Steve crosscheck a sample of your list vs. the campaigns own list to see what % of your list is fresh ?

I'm not sure if he can do that for me. I'll ask him.

BUSHLIED
07-08-2011, 11:04 AM
I think the Romney total tally or percent for group b is inaccurate, please double check...

"Results: Group B

There were 267 responses to the poll question as follows:

Barack Obama 101 (38%)
Michele Bachmann 29 (11%)
Sarah Palin 24 (9%)
Tim Pawlenty 13 (5%)
Ron Paul 12 (4%)
Mitt Romney 24 (3%)
Rick Perry 3 (1%)
Other/Undecided 61 (23%)"

Also, I am assume that because Group B are land-lines you can't determine gender.

rp08orbust
07-08-2011, 11:11 AM
I think the Romney total tally or percent for group b is inaccurate, please double check...

You're right, the percentage was incorrect.


Also, I am assume that because Group B are land-lines you can't determine gender.

Correct.

BUSHLIED
07-08-2011, 11:13 AM
It would also be helpful if you aggregated the samples based on gender, political status and age..leaving out zip code. So that you get a total sample size among all independents within this age range...

Aggregate Sample #1 (male, independent, below age of 40) Total votes for each candidate= Total sample size=

Aggregate Sample #2 (female, independent, below age of 40) Total votes for each candidate= Total sample size=

Thanks.

rp08orbust
07-08-2011, 11:23 AM
It would also be helpful if you aggregated the samples based on gender, political status and age..leaving out zip code. So that you get a total sample size among all independents within this age range...

Aggregate Sample #1 (male, independent, below age of 40) Total votes for each candidate= Total sample size=

Aggregate Sample #2 (female, independent, below age of 40) Total votes for each candidate= Total sample size=

Thanks.

It's complicated by the fact that different candidates have been listed in different polls. Also, different polling companies were used with differing response rates, so the results would have to be weighted.

cdc482
07-08-2011, 11:41 AM
Way to go dude! I will help financially, and please let us know if there's anything else we can do to help!

rp08orbust
07-08-2011, 11:43 AM
I'm not sure if he can do that for me. I'll ask him.

He said very roughly half of the supporters identified through robopolling were unknown to the campaign.

I expect that to increase when I ditch the voter database and adopt my new strategy. What we'll find out today is whether that is cost effective.

cdc482
07-08-2011, 11:43 AM
rp08orbust for Ron Paul supporter of the year!
Someone remind me how to send reps.

cdc482
07-08-2011, 11:46 AM
Am I right in this calculation?
7500 voters * $4/voter = $30,000 for Ron Paul to win Ames.

In fact, it is much lower than this, because every supporter identified can bring friends, and the cost will likely go down.

rp08orbust: Can you make a ticker for donations/supporters gotten? It would make this easier to spread around outside of RPF.

cdc482
07-08-2011, 11:46 AM
EDIT: Donated!

parocks
07-08-2011, 12:03 PM
Landlines 4% - cellphone females 6%. Looks like Landlines vs Cellphones is bigger than men vs women.

Excellent.

Potentially some hero stuff.

parocks
07-08-2011, 12:05 PM
what is "that" is cost effective - getting the list or calling the list for voter id?


He said very roughly half of the supporters identified through robopolling were unknown to the campaign.

I expect that to increase when I ditch the voter database and poll among the general cell phone population. What we'll find out today is whether that is cost effective.

BUSHLIED
07-08-2011, 02:12 PM
I should point out that while Ron's support among the "young" in Iowa is his largest bloc, they only represent roughly 11% of the entire voting block in Iowa's GOP primary. I would be curious to see the age group: 45-64 which represents the largest voting in IA according to CNN exit polls from 2008.

For example 11% of 21% (17-29yo) is much smaller than 9% of 46%(45-64).

If you take a theoretical population of 100 voters and break them down into age categories...then Ron get's 2.3 voters among the young and 4.14 voters out of the 45-64yo age group.

Based on this, one would hypothesize that there are more Ron Paul supporters among the 45-64 voting bloc. This IS the voting block that wins elections, plain and simple.

I would suggest polling the 45-64 year old voter segment that identify as independent. Use a mix of cell-phones and land-lines if you have to.

rp08orbust
07-08-2011, 02:32 PM
Based on this, one would hypothesize that there are more Ron Paul supporters among the 45-64 voting bloc. This IS the voting block that wins elections, plain and simple.

I would suggest polling the 45-64 year old voter segment that identify as independent. Use a mix of cell-phones and land-lines if you have to.

Very good points.

rp08orbust
07-08-2011, 02:34 PM
what is "that" is cost effective - getting the list or calling the list for voter id?

The list is free. It's the calling costs that will determine how cost effective this strategy is. I'll be testing it at 7pm Central.

rp08orbust
07-08-2011, 02:40 PM
Am I right in this calculation?
7500 voters * $4/voter = $30,000 for Ron Paul to win Ames.

This assumes there is an inexhaustible supply of phone numbers in Iowa to call that I haven't already called. If I'm identifying 22 Ron Paul supporters for every 13,500 phone numbers, as I did yesterday, then there would need to be 13,500 * 7500 / 22 = 4.6 million people to call in Iowa in order for me to find 7,500 Ron Paul supporters.


rp08orbust: Can you make a ticker for donations/supporters gotten? It would make this easier to spread around outside of RPF.

No I can't, unfortunately, but that's a great idea.

IndianaPolitico
07-08-2011, 03:17 PM
Keep up the great work! I enjoy coming on here and seeing the results of the robo polling.

trey4sports
07-08-2011, 03:29 PM
He said very roughly half of the supporters identified through robopolling were unknown to the campaign.

I expect that to increase when I ditch the voter database and poll among the general cell phone population. What we'll find out today is whether that is cost effective.

Ok, so based on you're above post were actually looking at $8 per supporter if 50% of the supporters identified have already been identified by the campaign. Even then, i'd be willing to bet once you expand into general, cell phone population we can inch that mark up to 65% or so unknown to the campaign, so lets round that number to a very conservative $6 per supporter identified.

So then lets assume 30% of those identified will go to the straw poll that becomes roughly $18 per straw-poll-going voter identified. Total cost would then be roughly $135,000 to identify 22,500 supporters with the assumption that 1/3 of those supporters will go to the straw poll equaling 7500 votes for Ron Paul that are directly attributed to the grassroots and are completely separate from the votes the campaign is bringing in.

So, looking at the numbers i have provided about are there enough calls to be made within a reasonable distance of Ames that could Identify the 22,500 Ron Paul supporters that would equate to 7500 straw poll votes? We would need to make somewhere around 10 million calls?

BUSHLIED
07-08-2011, 06:29 PM
I see this effort as "supplemental." Even if this effort resulted in an additional 500 Iowa straw poll attendees it would be worth the cost...

Keep in mind that in 2007, 14,302 ballots were cast. If you assume there was no voter fraud. Anyway, Ron got 1,305 votes. So even if attendance doubled and there were 30,000 votes casted, Ron would need 9,000 votes to win. The additional 500 voters would represent 5.5%, 1,000 would be 11%, and 2,000 would be 22%...all great considering the ease and cost of acquiring the names and numbers of these individuals.

Lets just say if the robocalls identified the original 2007 amount of 1,305 voters that would be a fantastic accomplishment.

I am a bit more conservative when it comes to amount of new supporters one is going to find and would be happy with a 5.5% success rate of 500 new voters.

This whole effort is based on the fact that the campaign is busing in Iowa residents to attend from within the state. That is a huge factor and I hope that everything is set up and there are no snags and that the effort is completely and throughly organized.

[I]If I were the campaign, I would have someone on each bus or someone counting each person coming off the buses when they arrive. The campaign should know how many people they bused in total. There should also be a sign in for those not having bused in but walked or drove in. You get my point. With Obama on the ropes and the GOP split, the corruption within the GOP is going to be at an all-time high This is a high-stakes election considering how all the issues we are facing play well for conservatives and a chance to pick-off an incumbent with a socialist agenda. Should Obama lose, the partying in the streets of Red America would last for weeks!!

HarryBrowneLives
07-08-2011, 09:22 PM
[QUOTE=rp08orbust;3386006]This assumes there is an inexhaustible supply of phone numbers in Iowa to call that I haven't already called. If I'm identifying 22 Ron Paul supporters for every 13,500 phone numbers, as I did yesterday, then there would need to be 13,500 * 7500 / 22 = 4.6 million people to call in Iowa in order for me to find 7,500 Ron Paul supporters.


Did a little homework for ya ... According to the 2010 Census there are 2,324,368 people in IA over the age of 18. So if we devide that number by 13,500 times 22 =
A hypothetical 3,787 potential Ron supporter we could find via Robo. Not bad at all. :)

This is also my 1,000 post. Kwool!

HarryBrowneLives
07-08-2011, 09:50 PM
We should create a Sales Funnel approach with this (that's what I do for a living) ...

Supporters IDed
Supporters who want to vote (or register) to do so for RP
Those same supporters Robo'ed (later) for the moneybomb
Supporters who would like ride to Ames
etc, etc.

If this works in Ames, the ROI for the campaign on this could be super. I think the key is to get people to donate something ... anything. They will be more likely to be regular donations and voters with a further vested, emotional, interest. Dunno?

Lotta ideas here that, if work, could be a real way to be effective pretty much everywhere. The edge we have here in this race is hardcore, skilled, activists who can put boots on the ground coupled with the fact that RP's numbers are something largely unseen in modern politics ... we're not big (yet) , but we are Krazy Glue with attachment ... our voters are a mile deep while others are a mile wide. That is the test is how effecient we can be. If we can GOTV on steroids we will be tough anywhere. If you don't have mile wide support you have to have turnout. General apathy, rainy days, etc. are our friends:)

harikaried
07-08-2011, 11:28 PM
I wonder if it'll be effective to poll past democratic voters asking about the wars in Libya, PATROIT Act, TSA as per http://www.huffingtonpost.com/robin-koerner/blue-republican_b_886650.html

If you're looking for a numeric response, perhaps press the number corresponding to the # of topics agreed/disagreed.