|
Title: Fivethirtyeight: Are bad pollsters copying good pollsters? Post by: Mr. Morden on August 12, 2014, 10:46:52 AM http://fivethirtyeight.com/features/are-bad-pollsters-copying-good-pollsters/
Quote So, why do the nontraditional pollsters seem to do worse in races where there aren’t gold-standard pollsters conducting polls? The above chart looks similar to one produced by Princeton University graduate student Steven Rogers and Vanderbilt University professor of political science Joshua Clinton, who studied interactive voice response (IVR) surveys in the 2012 Republican presidential primary. (IVR pollsters are in our nontraditional group.) Rogers and Clinton found that IVR pollsters were about as accurate as live-interview pollsters in races where live-interview pollsters surveyed the electorate. IVR pollsters were considerably less accurate when no live-interview poll was conducted. This effect held true even when controlling for a slew of different variables. Rogers and Clinton suggested that the IVR pollsters were taking a “cue” from the live pollsters in order to appear more accurate. My own analysis hints at the same possibility. Here's the damning graph: () Title: Re: Fivethirtyeight: Are bad pollsters copying good pollsters? Post by: SPC on August 12, 2014, 04:55:29 PM There are better explanations than mere cheating though. It is also possible that such pollsters choose not to release polls when they deviate significantly from the consensus of good pollsters, thus making the polls they do release in such states more accurate as a whole.
Title: Re: Fivethirtyeight: Are bad pollsters copying good pollsters? Post by: Mr. Morden on August 13, 2014, 08:22:29 AM There are better explanations than mere cheating though. It is also possible that such pollsters choose not to release polls when they deviate significantly from the consensus of good pollsters, thus making the polls they do release in such states more accurate as a whole. That's cheating too. It means you're covertly throwing out data points that disagree with other people's results so as to make your overall approach appear more accurate than it already is. Title: Re: Fivethirtyeight: Are bad pollsters copying good pollsters? Post by: SPC on August 13, 2014, 12:08:17 PM There are better explanations than mere cheating though. It is also possible that such pollsters choose not to release polls when they deviate significantly from the consensus of good pollsters, thus making the polls they do release in such states more accurate as a whole. That's cheating too. It means you're covertly throwing out data points that disagree with other people's results so as to make your overall approach appear more accurate than it already is. Sorry, I should have phrased that better. I meant that moving the goalposts is a more likely explanation than mere R2K/Strategic Vision style pulling numbers out of thin air. |