Saturday, June 25, 2016

One form of polling was dead accurate on Brexit, and it’s changing how we predict elections

One form of polling was dead accurate on Brexit, and it’s changing how we predict elections




Poll after poll predicted the Remain vote would win the day and there would be no Brexit. Almost all of them got it wrong. Thursday’s referendum left the UK preparing to leave the European Union, and many pollsters wondering what went wrong with their efforts to gauge public sentiment.

The polling industry’s crisis of confidence is fueled by several recent failures to forecast major outcomes. From British Conservatives’ victory over Labour in 2015 to Trump’s unexpected dominance of the Republican presidential primary race (you can read celebrity pollster Nate Silver’s 5,379-word mea culpa on that here), well-respected polling outfits seem to keep getting it wrong.

It’s likely an issue of changing technology and communication behaviors. The former “gold standard” polling technique—randomized phone calls to a representative cross-section of the population—is increasingly unreliable, given the spread of mobile phones, government bans on automated calling technology, and declining response rates,reports The Huffington Post’s polling editor, Natalie Jackson. After a statistical review of the three most common polling methods—live telephone, internet, and automated phone—she wrote, “we can’t really predict which type of polling method will be most reliable in pre-election surveys”

One survey method, however, has proved more accurate than the rest, and is promising to become the future of polling: internet and mobile phone polling. While not perfect, it gives surveyors the insights of personal interviews with the accuracy and efficiency of modern technology. By reaching out through apps, phones, and websites—and collecting a suite of ancillary data—surveyors are finding they can consistently deliver more accurate data.
The Real Style ------- NP 2016

No comments:

Post a Comment