I had always made the assumption that there must be a reason for someone being elected. So, if Trump won over Clinton then that surely must have meant that he had strong support, nearly half the voters. But then I watched a youtube video that someone in another thread linked to where a man who was a mathematician talked about statistical sampling. And he brought up something that had not occurred to me.
He talked about doing a survey for something like Coke vs Pepsi. He said that if there was a strong preference that you would expect this to be shown in the data. Yes, that makes sense. But then he brought up the fact that if there was no strong preference, if most people viewed Coke and Pepsi as about the same then you could still get data with many more votes for one or the other but that the data would be meaningless. That was something that I hadn't thought about.
That brings up a question that I don't recall being addressed. If people saw Clinton and Trump as roughly equal as candidates then Trump's victory could just be random. It might carry no meaning at all. I've tried to figure out if there was any way to tell if this was the case.
It could help explain the disparity between polling and election results.
It could explain why Trump's approval decline has surpassed that of even Gerald Ford which was itself, unprecedented.
Are there polls, attitudes, or behaviors that anyone can think of that could prove or disprove this?
He talked about doing a survey for something like Coke vs Pepsi. He said that if there was a strong preference that you would expect this to be shown in the data. Yes, that makes sense. But then he brought up the fact that if there was no strong preference, if most people viewed Coke and Pepsi as about the same then you could still get data with many more votes for one or the other but that the data would be meaningless. That was something that I hadn't thought about.
That brings up a question that I don't recall being addressed. If people saw Clinton and Trump as roughly equal as candidates then Trump's victory could just be random. It might carry no meaning at all. I've tried to figure out if there was any way to tell if this was the case.
It could help explain the disparity between polling and election results.
It could explain why Trump's approval decline has surpassed that of even Gerald Ford which was itself, unprecedented.
Are there polls, attitudes, or behaviors that anyone can think of that could prove or disprove this?
via International Skeptics Forum http://ift.tt/2gqZmxH
Aucun commentaire:
Enregistrer un commentaire