Showing posts with label Matthew MacWilliams. Show all posts
Showing posts with label Matthew MacWilliams. Show all posts

Wednesday, January 20, 2016

That troubling Trump supporter as "authoritarian" poll you're hearing about today - More proof that U.S. presidential polling is biased, unreliable and full of ideoological traps designed to prove... "something." But showing something once in a poll is not PROOF, just a one-time result. Reliable polling is getting the same/similar results over and over consistently thru objective means





The One Weird Trait That Predicts Whether You’re a Trump Supporter 

And it’s not gender, age, income, race or religion.

By Matthew MacWilliams
1/17/2016
http://www.politico.com/magazine/story/2016/01/donald-trump-2016-authoritarian-213533
It's not so surprising that such an ideological survey would first appear in Politico.
But it naturally leads to the questions not asked or mentioned, like...well...
Question: What's the one trait that predicts whether you're a Hillary supporter?
Answer: They are NOT interested in that answer.

(Though it once was, "And in the 2008 Democratic primary, the political scientist Marc Hetherington found that authoritarianism mattered more than income, ideology, gender, age and education in predicting whether voters preferred Hillary Clinton over Barack Obama.")
 
Politico is no more interested in revealing that than they are in knowing and publicly disclosing whatever the supposed one magic trait about supporters of Bernie Sanders is.
Especially if that were to be something like, oh, people who despite saying very liberal and progressive things in front of strangers and the news media, when it comes down to it, do NOT want to live near people who are similar to them, which is the most plausible answer one can infer from facts like Sanders' support being strongest in almost entirely 100% White enclaves around the country, something Hillary is currently exploiting in South Carolina with its large Black population.

In case you did not know, the only state in the U.S. with a lower percentage population of minorities than the state Sanders represents, Vermont, is Maine.
I know because I checked it out via the latest census info a few months ago, and even found similar numbers on ye olde Wikipedia, though the latter says Montana instead of Maine, two states that could not otherwise be more dissimilar from one another.
https://en.wikipedia.org/wiki/List_of_U.S._states_by_African-American_population

It doesn't bother me, per se, that people with particular biases have opinions and want to sound off on them, since everyone is free to believe whatever they want, however crazy or different from my own POV it might be. After all, it's a long campaign...
But what I hate seeing and find troublesome is the way this story is already being played up nationally as evidence of... well, "something."
But all it really is is a SINGLE snapshot in time.

It's like predicting the Miami Dolphins making the NFL playoffs every year based on them frequently beating the Patriots the past few years when they have been one of the best NFL teams. 
But in those years when they do beat the Patriots -almost always at home- they STILL fail to make the playoffs, don't they? (This year proved that all over again!)
 
Experienced football fans who have some real knowledge and historical perspective, like political junkies with the same qualities, know that one result is often an outlier. 
What you need to see is consistency (of effort) and results.  
Results plural.

Right now it's a theory that will not be PROVEN until it can be successfully replicated in multiple objective polls. And the article doesn't even have any links to check the poll numbers and questions yourself.  
WTF?

We seem to be at the point where someone who wants the public to believe something in particular about a candidate, and try to be seen as above reproach, and merely relying on cold hard numbers, can write something ascribing far-reaching significance...after just one poll.
Well, it doesn't seem like anything resembling polling Best Practices to me.

And now that you think about it, if this sort of designed poll is such a great thing, how come we never heard from the U.S. news media about the results of the same designed poll in 2008 and what it supposedly "said" about Hillary's supporters?
Why am I only hearing about it eight years later?
Here's more irony. T
oday, hours after seeing lots of tweets about the above, I saw this:




Posted by at No comments:
Labels: , , , , , ,
Subscribe to: Posts (Atom)