Transparency, Polls & Surveys, Episode 103
By Brenda Ann Eckels, aMGC (c)2016
It started with Shell Shocked Working Class Whites on CNN. The story led to conclusions that did not reflect any of the white working class people I knew except my rabidly Tea Party GOP Trump loving brother. I smelled something.
That led to the actual survey report that CNN & Kaiser paid for and did just in time, coincidentally to make front and center on today’s news, even though it was put out in September. A decidedly pro-Trump barrage of stories and video on the same day as the Trump Tape went public. Hmmm. My sinuses started to fill up with mucus.
I have some issues with this survey, not the least of which is that CNN spun no less than 4 separate stories and op-eds plus a video with cool flashy red themed graphics and pushed it at least twice…
….on results from exactly 701 working class white people they “carefully surveyed”. An entire series of blog posts and questions were born. Let’s start with:
Since When Do 701 Working Class Whites
Get To Dictate What
ALL The Millions Of
Working Class White People Think or Vote For?
701 opinions is so infinitely small that it can not possibly be reflective of even the members of the Log Cabin Republicans, never mind millions of humans from McDonald’s workers to the folks at the nearby wind turbine factory!
That’s the Focus of This Episode, Number 103, of Transparency, Polls, & Surveys!
Who were these people? Where did they come from? Were they picked from a list, like maybe registered voter lists? Or some random telemarketer’s list? From a Twitter Follower List? Who’s list?
Oh, no not a list, an RDD.
What’s an RDD?
Random Digit Dial.
Yep, the computer just randomly mixes up numbers and calls, and then asks the person who picks up if they are over 18 and if they live in America.
The computer did this – in this case – 1,614 times.
- Yes, the computer randomly mixed up numbers.
- Then it sorted the numbers into landlines and cell phones.
- For the landlines they had the computer think of a random month between January and December and then dial the number.
The first question to landlines was:
“Can I speak with the person in the household who’s birthday is closest to (insert whatever month the computer had just thought up) please?”
Then the computer asked if that person was over 18 and if they lived in America.
Cellphones were easier. They just asked whoever answered the phone if they were over 18 and lived in the US. (In my home, this is usually my home care aide.)
The computer mixed up numbers and called 1098 numbers.
Not surprisingly in a very diverse country
with an estimated 2 billion humans eligible to vote,
they did not get enough working class white people,
which is what they were looking for because:
graphics arts at CNN already had started working
on this cool red themed set of graphics,
and one writer had his op ed half finished.
What happened next? This is not a transcript, I am paraphrasing here….
Scene: Big room, one half CNN staffers, one half Kaiser staffers. Stale Coffee. One really big computer.
Uhh, Computer: Go find the US Census results and pick the census blocks with large numbers of white people with no 4 year college degree.
Got it? Good. Now mix up random numbers just in those areas, and call.
(They found 294 more people, but only 181 were white working class.)
Uhhh, we need more white working class.
Computer: Go back to all our other surveys where we asked if the person was over 18 and white and had no college degree at all. Got it? Yeah, go ahead and pick out 77 of those guys.
Oops, now we are short 52 blacks. We have to make the total number the same percentages as the US population.
Computer, Go back to all our other surveys where we asked if the person was over 18 and black and had no college degree at all. Got it? OK, pick 52 of them.
Crap. We forgot the Hispanics again. Why do we always do that?
Computer: Go back to all our other surveys where we asked if the person was over 18 and Hispanic and had no college degree at all. Got it?
Deep bass voice of computer interrupts
“Excuse me Sirs, but research has made very clear that most Latinos do not like being called Hispanic…”
Yeah, whatever. Look it doesn’t matter what we call them if Trump wins because there will be a million less of them. Just…Just go pick 35 of the Hispanic Latino wetbacks what-ever. C’Mon Computer, We need to get this on the air!
That is exactly what is condensed down to a partial paragraph of the very tiny print that is the most important part of any survey or poll:
The Methodology Statement.
Always read the Methodology Statement first, no matter how cool the red flashy graphics are or how many stories a tv network spins out of one set of data. You can click here and read the survey methodology for yourself to fact check me on how many calls and how the grand total of 701 white working class people got to pick the winner of the election early for about 30 minutes….
….before said “winner” shot himself in the foot with a lame video apology for being a bigoted pig. Most of America already knew this about him, so both his 2005 vulgarities and his new lame apology got a lot of play at #slownewsday. His supporters, including Rachel Pendergraft — the national organizer for the Knights Party Ku Klux Klan and those over at the only other party to endorse him, have his back, saying they say the same stuff.
Next episodes: How many people lied in the phone call? Were ANY of them registered voters? How many belong to the KKK? Why did it take from August 9 to September 5 to make less than 2,000 phone calls? How long did CNN know about the results, and how did it affect the OTHER stories they did before this one? Who paid what? Who owns Kaiser? CNN? Why did Stephanie Coontz have to update her story September 23rd? What did she change? Will CNN Graphics Designers pay more or less taxes under Gary Johnson? and more!