We’ve written at length about the rapid changes in the ways in which Americans communicate and the impacts on political professions, particularly survey researchers. But Tuesday brought two fresh examples of the way the industry is reacting to those changes — examples that represent a stark departure from previous polling conventions.
— Public Policy Polling, the prolific Democratic robopollster, announced Tuesday that 20% of their interviews will now be conducted via Internet, in an effort to reach the more-than 40% of Americans without landline phones. We’ve long been critical of PPP for excluding cell-phone-only voters, and, despite the usual caveats about opt-in Internet polls (and others about mixed-mode surveys combining more than one methodology), this seems like a step in the right direction.
— The other (more significant) story comes from Glen Bolger and Trip Mullen at Alexandria, Va.-based GOP polling shop Public Opinion Strategies. In a story for Campaigns & Elections, Bolger and Mullen outline the experiments POS has been conducting with mobile surveys — that is, polls completed via mobile app, which they say “offers an interactive experience for the respondent that’s not possible over the phone or even through online research. The application allows survey researchers to harness mobile device capabilities like touch screens, built-in cameras, and GPS positioning.” (The most entertaining part of Bolger and Mullen’s piece: The photos voters send when asked to take a snapshot on their phones of something that reminded them of the two political parties!)
— While POS and other firms are starting to do lots of work online, this work is viewed more as a supplement to their increasingly-expensive, call-based phone polls. In other words, the horse-race telephone poll isn’t going anywhere, even as Americans increasingly replace landlines with cell phones (and replace cell phones with smartphones that support these kinds of apps). “I see them as different products with different objectives. I don’t see how, today, from the work we’ve done in 2013, they oughta be blunted into one survey response,” POS’s Bill McInturff said last year. “The political pollsters will be the last, last, last people on the phones.”
Web- and mobile app-based polling will always be anathema to some in the survey research community because they eschew the principle of probability sampling — the idea that every member of the sampling frame (registered voters in Virginia, for example) has an equal chance of being selected to participate. But every household doesn’t have a home phone anymore, and calling cell phones is — in some cases — prohibitively expensive. It’s clear that non-probability web and mobile research is progressing — on two tracks. Public and media polls are using opt-in Internet and mobile surveys to replace more expensive, live-caller efforts. PPP joins Rasmussen Reports and SurveyUSA in doing cell-only subsamples via the Internet, and the Associated Press and Reuters have already moved entirely online. On the campaign side, Internet and mobile polling is increasingly a major tool for ad and message testing and other supplemental research, but the basic horse-race, brushfire poll model remains the dominant mode, despite rising costs.
What We're Following See More »
"Members of the Congressional Black Caucus are reviving calls to remove Confederate statues from the Capitol following the violence at a white nationalist rally in Virginia." Rep. Cedric Richmond, the group's chair, told ABC News that "we will never solve America's race problem if we continue to honor traitors who fought against the United States." And Mississippi Rep. Bennie Thompson said, “Confederate memorabilia have no place in this country and especially not in the United States Capitol." But a CBC spokesperson said no formal legislative effort is afoot.