SoonerPoll.com, Oklahoma’s public opinion pollster, has announced that it will be changing its methodology.
SoonerPoll.com’s methodology will now include a combination of landline and cell phone random digit dial samples, known as a dual-frame sample.
“The world is changing, and we want to continue to make sure that our sampling remains representative of all voters likely to vote on Election Day,” Bill Shapard, CEO of SoonerPoll.com, said.
Based in Oklahoma City, SoonerPoll performs its own data collection with on-site calling centers, which produce high quality data with strong internal validity.
Prior to this, SoonerPoll has used an internal database of registered voters, or a list-based methodology. In 2004, SoonerPoll developed a statistical model that identified likely voters by evaluating past voting history and other factors to create a likely voter sample. Participants were then randomly selected.
Shapard said that only about 65% percent of likely voters in Oklahoma can be reached via a landline phone today, compared to 94% in 2004.
“We have talked before about how using a list-based methodology to quantify likely voters increases the likelihood that a phone match can be found for participants in a random sample,” Shapard said. “But, as more and more families become cell phone only households, the writing on the wall says this kind of methodology will soon be unsustainable.”
Shapard went on to say that SoonerPoll’s change in methodology will ensure that it stays ahead of the curve on the cell phone issue, which continues to put public opinion experts around the world in a bit of a predicament.
While SoonerPoll is moving away from the list-based sampling model, it has had great success with it in the past. SoonerPoll has been extremely accurate, boasting an error rate smaller than all other pollsters in Oklahoma’s last four general elections.
Sampling both landlines and cell phones is a technique used in traditional market research when trying to sample the adult population, but there is still the matter of identifying likely voters.
Once a random sample is selected, participants will be subjected to SoonerPoll.com’s comprehensive likely voter screen. SoonerPoll.com’s likely voter screen mimics Gallup Poll’s screen and ensures that the people SoonerPoll talks to are the same people you would expect to see at the polling place.
After all of the interviews are conducted, the combined landline and cell phone samples are weighted based on several key factors, including landline and cell phone data obtained through comprehensive health surveys conducted by the federal government. Demographic data, such as sex, age, political party, and congressional district are also used in the weighting.
When new polling results are released, SoonerPoll will also supply a comprehensive Sample Disposition and Rate Calculations report. This report with include a disposition of all calls made from the sample and calculate response, cooperation, refusal, and contact rates as defined by the American Association for Public Opinion Research (AAPOR). This new disclosure is part of an industry effort to provide more disclosure of how polls are performed, known as the AAPOR Transparency Inititiative, of which SoonerPoll is active participant.
“With this new disclosure, SoonerPoll will be the first Oklahoma pollster to release such data and will be providing more transparency of how its polls are performed — more than any other polling firm in the state,” Shapard said.
SoonerPoll.com will purchase its samples from Survey Sampling International (SSI), the world’s leading provider of sampling solutions for survey research.
SSI uses a variety of random digit methodologies to address landline representation issues. SSI ‘s wireless/mobile sample reduces the potential of coverage bias by providing a non-overlapping sampling frame that complements their Random Digit Dial telephone sample. SSI has made more information about their wireless methodology available.
More information of SoonerPoll’s new and previous methodology can be found at SoonerPoll.com.