How Bad Polling Will Normalize Nazism
According to this CNN/ORC poll, CNN claims that the majority of all Americans actually think Trump is doing a good job:
A narrow majority (53%), say they think Trump will do a very or fairly good job as president, and 40% say they have a lot of confidence in Trump to deal with the economy, a share that outpaces the percentage who had that much confidence in Barack Obama, George W. Bush, Bill Clinton or Ronald Reagan ahead of their first inaugurations.
Except let me, a trained sociologist, tell you how polling lies.
Let’s start with the methodology.
This poll called 1003 people over the course of three days, half of them on landlines. LANDLINES. Do you know anyone under 40 who owns a landline? How many working people do you know still maintain a landline? Spoiler alert: most American’s don’t.
In fact, so few Americans have landlines anymore that it’s causing huge problems for emergency responders.
Now ask yourself seriously how many rational people you know would answer a 21-question poll when an unknown number calls — keeping in mind that after collecting basic demographic data, this survey would take at least 20 minutes to complete. Nothing in the methodology says participants are reimbursed for their time or incentivized to participate in any way.
Right, as you suspect, what happens with phone polling then is that the vast majority of people refuse to participate (notice how they never tell you how many decline participation — bad survey methodology).
So by virtue of relying on phone surveys for polling, you’re trying to take the pulse of America by sampling on exclusively statistical outliers. That’s the *weirdest* kind of sampling bias.
Think about the profile of someone who is willing to sit down and do this survey. They hear it’s for CNN and they get excited because they probably sit at home and watch a lot of CNN. They are the kind of passive person who thinks this is all very exciting. In other words, they are probably very basic.
And by sampling on basic, you make America seem like the stupidest fucking country on Earth — time and time again.
But because most people only read the headline, much of this scrutiny just gets buried. There’s no time to unpack it. There’s no time to make sense of how useless this kind of polling is because right-quick we’re on to the next poll and its absurd findings. Would you have stopped to question the methods of the polling behind headline? Be honest.
This election taught cable news that the facts don’t matter. You can throw up any numbers you want and people will read them like tea leaves. We’re in a post-truth age. And as I’ve said before, numbers are numbing.