PoliMetrics

 

 

Here's How PoliMetrics Works In 5 Easy Steps:

Build a Contrasting Survey

We will work with the campaign to identify 2 or 3 likely issue contrasts that will separate the candidates.  We then build a survey instrument of 6 to 7 questions that begins with an initial, uninformed vote for candidate X or candidate Y.  The next few questions will be issue comparisons meant to provide contrasting messages that are written to slightly skew toward your candidate but are a fair representation of the opponent's position on key subjects.  Finally, we ask the informed vote to determine the movement among voters from the initial vote to the informed vote. 

A survey like this will provide us a number of modeling opportunities.  From the initial vote we will be able to predict the likely GOTV audience.  Each issue comparison will provide us data to model the most effective audience for each of the issues.  And the movement from the initial vote to the informed vote will allow us to model the overall persuasion universe. 

Use Your Volunteers To Collect The Data

Most other modeling companies use IVR or robo polling technology because paid callers are very expensive when our goal is to ultimately obtain thousands of completed surveys. 

There are problems with using IVR.  For one, you don't know with certainty that the data you're collecting is for the specific person on the file.  Even if only a small percentage of surveys are attributed to the wrong voters, it will have a negative effect on the accuracy of the model.  We need the model to be as accurate as possible.

Secondly, an IVR survey that poses more complex issue descriptions does not allow for the voter to hear the question again or get clarification.  This, too, will impact the accuracy of the model.

In our view, PoliMetrics is a field organizing tool that compels the campaign to begin building a robust volunteer operation well in advance of election day. 

Using our Activate predictive dialing technology you can and should use volunteers to conduct the surveys.  This provides your field operation a significant need to build your volunteer operation and avoids the pitfalls of IVR programs.

Four Phases Of Statistically Accurate Sampling

We recommend conducting the project in four phases of 800 completed surveys.  We will work with the campaign to pull the sample in the same way a polling firm pulls a sample that is random, yet representative of the district.  We will work with you to monitor the calling and provide cross tabs so that we are able to avoid harmful oversampling of certain demographics by adjusting the data to accommodate more accurate results.

While we do not recommend using PoliMetrics as an inexpensive replacement for polling, by running the program in phases you will have statistically accurate data during each phase so you can gage support and track improvement or erosion.

Build The Model And Score Your Voter File

After the first 2 phases we can begin the regression analysis and will have a phase 2 score for your voter file within a few days of completion.  Each phase will improve the model.  All 4 phases must be done by late September or early October, depending when early voting begins.

Our modeling capability is cutting edge using the most sophisticated statistical analytics package  and includes a double blind control group with each phase to test the efficiency of the modeling scores.

Save Money

The final cost of a PoliMetrics program will vary based on the number of surveys collected and the length of the survey.  The cost for the program described here would cost about $12,500.00.  We encourage you to compare this cost with other modeling companies pricing.  We know it's very competitive.