With the next US presidential election still two and some years away, we're OK with poking some fun at the hopefuls' expense... like the Washington Post did in last week's Wonkblog piece that compared how the politicos are polling against Star Wars characters. The Force, we've learned, is not with them.
Even Hillary Clinton, who with a 19% net favorability rating bested the bunch of the 2016 White House wannabes but only scored as well as bounty hunter Boba Fett, found fun in the inanity -- because pure silliness is what it was. Data journalist Christopher Ingraham combined legit FiveThirtyEight results from a poll on favorite Star Wars characters with his assessment of the current candidate field, determining that not a one of the 2016 hopefuls rank higher in people's minds than the formidable Darth Vader.
During a CNN interview that aired Sunday, Clinton said that, what with Darth Vader being so dark, he wouldn't be her choice. Rather, she said, she'd select "somebody
of perhaps a slightly more positive attitude in his presentation."
That we can laugh at the Wonkblog's farcical poll results is in part because we know that there's real, time-tested science that goes into official presidential polling. Come the summer of 2016, prognosticators will be opining on who'll land in the Oval Office because of what the polls tell them. Pollsters will be taking the pulse of the American public -- maybe even you or me -- through the time-honored tradition of representative sampling.
Representative sampling may have survived the test of time, but some researchers find that it's now time to let it rest. Its effectiveness and value is diminishing, researchers at Columbia University and Microsoft Research wrote in a paper, "Forecasting Elections with Non-Representative Polls," published this spring. They say polls are too costly, too time-consuming, and too outdated a methodology for today's always-connected, electronics-oriented population. They note declining response rates and suggest the "statistical benefits of representative sampling have diminished" as well.
Of course, they have an alternative -- one that might at first sound as ludicrous as comparing how 2016 hopefuls are fairing against Han Solo and his gang. Let's just ask Xbox gamers for which candidate they're going to cast their votes!
In the paper, the researchers show "that with proper statistical adjustment, non-representative polls can be used to generate accurate election forecasts, and often faster and at less expense than traditional survey methods." They demonstrated how by creating 2012 presidential election forecasts from the results of daily voter intention polls taken by Xbox players -- "a novel and highly non-representative survey dataset." They continued:
After adjusting the Xbox responses via multilevel regression and
poststratication, we obtain estimates in line with forecasts from leading poll analysts, which were based on aggregating hundreds of traditional polls conducted during the election cycle. We conclude by arguing that non-representative polling shows promise not only for election forecasting, but also for measuring public opinion on a broad range of social, economic and cultural issues.
As they explain in the paper, the researchers made an opt-in poll, with three to five questions, continuously available to Xbox players during the 45 days preceding the 2012 US presidential election. First-time respondents were asked to provide basic demographics, as well as political ideology, and the name of the candidate for which they voted in 2008; players could only participate in the poll once daily.
During the 45 days, the researchers conducted more than 750,000 interviews with almost 346,000 unique respondents, more than 30,000 of whom completed five or more polls. This base, they note, made this one of the largest-ever election panel studies -- and so who's to say that Xbox gamers shouldn't share their opinions and help us figure out what we, the American people, are thinking about the contenders as they duke it out in the 2016 race? Or maybe pollsters should just keep trying to get somebody other than white-haired old ladies to pick up the telephone and answer a few questions.
Do you think we ought to use alternative polling during the 2016 US presidential campaign? If not by polling Xbox gamers, then how?
— Beth Schultz, , Editor in Chief, AllAnalytics.com