by Madeleine KandoLike many of you, I have been trying to make sense of the Trump phenomenon. How did we get to the point where almost half of the American voting public considers electing a buffoon as our next Commander in Chief?
There are many theories out there trying to explain this: 1. He speaks to the anger of many Americans who are disillusioned politically and suffer economically. 2. He fills a power vacuum in the Republican Party. 3.The sheer volume of media coverage has turned into a self-perpetuating cycle. Etc.
It is also possible that Trump’s incredible power of self-promotion has created the dreaded ‘feedback loop’. Political scientists point to a strong correlation between media coverage and a rise in the polls, especially early on in a campaign. And being chosen to participate in the primary debates also depends on where you stand in the polls. In other words, one step rests on the previous one, which has resulted in Trump’s mystifying success.
It also helps that he talks about every issue in vague terms, leaving room for interpretation of what he says to fit many foot sizes. His BS is an intentional maneuver to disorient and take the sails out of his opponents’ rebukes. He is a master strategist. Maybe it is true that he knows more about the military than the generals.
Unfortunately, the same feedback loop has also caused Hillary to get trapped in a negative spiral. Negative stories about her have been talked about over and over again, ad nauseam. Trump’s response to accusations is to not admit guilt, ignore the accusation and move on to something else.
The real problem with polling is that the results of the polls in one area affect people’s opinion in another part of the country. This is terrible. You get either a self-fulfilling or a self-defeating prophecy. In psychology they call it the ‘bandwagon effect’, when people do something primarily because other people are doing it, especially when they are uninformed.
Is it a coincidence that Trump’s current campaign manager is a pollling expert? Manipulating the ‘will of the people’ is Trump’s middle name, and he is incredibly good at it. He knows that polling not only measures public opinion, but also SHAPES it. leave comment here
1 comment:
Hi Mado,
It’s a bit late for this reaction - it’s 2 days before the presidential election, and “the cake is baked,” as they say.
But since I wrote this reaction to you back in September, when you wrote your piece on polling, I might as well post it (survey research was a huge part of what we had to study, in a Sociology graduate program, so whenever the subject comes up, I feel compelled to comment...)
By now, polling methodology can be very good . When done right (e.g. like Nate Silver’s Five ThrirtyEight), it’s a true science and its predictions can be extremely true.Statisticians have become very sophisticated at reducing the chance of error.
Of course there are many bad polls (e.g. those commissioned by biased parties like Fox News). But good professional polls like Gallup are rarely wrong. The last spectacular prediction error in a presidential race was in the 1940s, when Truman ran.
The “margin of error” means that a prediction is being made either with a 1%, a 5%, or some other probability of error. In other words, a poll prediction may be made, with the qualifier, saying: We are 99% sure, or 95% sure that we are correct in our prediction. These are also called “confidence intervals.”
As you said, the larger a sample is, the better your chances are of being correct (other things being equal of course – the sample should also be representative, etc.). So you can only be 100% sure if you interview the entire population. But if you interview 1500- 2000 people (which is what Gallup usually does), you have a high probability of being correct (at least 99%).
You are right: The real problem is that the results of polls AFFECT the way people vote. Election results are sometimes already predicted on the basis of results back east, when people in California and Hawaii haven’t even voted yet! So you get either a self-fulfilling or a self-defeating prophecy (people out West don’t bother to vote, since the outcome has already been called).
Now there are rules/laws to prohibit this, I believe.
It’s good that you mention the bandwagon effect. And you are totally right that polling is a bad thing in that it not only measures public opinion, but also SHAPES it.
When done right, polling results are usually pretty much a TRUE measurement of people’s attitudes AT THAT MOMENT.
Post a Comment
Please limit your comment to 300 words at the most!