Wednesday, November 2, 2016

Is Polling Just an Exercise in Chasing What We Already Knew? Maybe.

Sometimes we read a poll that tells us what we didn't know, and it can show us trends (more people aren't afraid of the gay than before), but in a political season, "response rate" and "reversion to the mean" show us that we're chasing voodoo already baked into the cake.


Kevin Drum shows us the futility of "hoping for a game changer" when the game was already won or lost, depending on who you've been rooting for all along:
...Although we didn’t find much vote switching, we did notice a different type of change: the willingness of Clinton and Trump supporters to participate in our polls varied by a significant amount depending upon what was happening at the time of the poll: when things are going badly for a candidate, their supporters tend to stop participating in polls. For example, after the release of the Access Hollywood video, Trump supporters were four percent less likely than Clinton supporters to participate in our poll. The same phenomenon occurred this weekend for Clinton supporters after the announcement of the FBI investigation.
So polls zig-zag, and then revert to the mean:
Up through July, Hillary Clinton was ahead of Trump by a steady 3.5 points. Then she got a big spike after the Democratic convention, but it quickly reverted to around 3.5. Then she dropped a bunch after some email news and her fainting spell on 9/11—but again, the trendline quickly reverted to 3.5. Then she spiked again after the second debate and the Access Hollywood video, but reverted to 3.5 yet again. 
Looked at this way, all our checking FiveThirtyEight, HuffPost Pollster, and RealClearPolitics has been for naught. "Ah, Hillary's better now!" was just statistically anticipated phenomenon.

Oh well, it was fun, anyway. Who's ahead now?!?


No comments:

Post a Comment