Monday, July 15, 2013

Polling on Zimmerman

I don't want to spent a lot of time on the Zimmerman trial.  Actually, I don't want to spend any time on the Zimmerman trial, but let's take a quick look at the polls after the not guilty verdict.

Frankly, most of the polls suck.  They're not even polls, at least in the real sense.

They're SLOPs, self-selected polls, which is a scientific way of saying they're complete bullshit.  Why?  Because the only people participating are those who happen to go to that site and care enough, or are pissed enough, to participate.  A quality survey uses a random sample, meaning everyone theoretically has a chance of being included.

This is an example from of all things a public radio station out of California.  I did it, and about 52 percent go with Zimmerman and self defense.  This TV station poll is no better, and with over 70 percent favoring Zimmerman.  And then there are bad polls, like this awful HuffPo (the results, interestingly, are about one-third for each of the response alternatives).  Hey, and lawyers get into it to, this poll from a lawyer web site (most thought Zimmerman not guilty).  Finally, the Orlando Sentinel, the nearest big paper, got into the act and also slopped its way to a 70 percent finding for Zimmerman.

There's nothing wrong with these polls -- as long as you label then non-scientific and completely useless and less a measure of public opinion than a way to engage our audience.  Simply put, for news sites, they're misleading.  They make people think they measure public opinion when, instead, they measure a handful of opinions of those who happen to visit the site and happen to participate and probably are skewed one way or the other in the first place.  In simpler, more academic terms, they're bullshit.

Okay, Hollander, what about the real polls about the verdict.  What do they say?

I'm waiting to see one, and expect a few out later today.  If I had to guess, I'd say between 55 - 65 percent favor the verdict, but I'm lousy at predicting public opinion.

No comments: