ABOUT POLLING
Back in 1972, the Democrats did not pick an ideal candidate with George McGovern. As the last votes were counted, Senator McGovern had about 29 million votes. His opposition President Richard Nixon had 47 million. So the President won. Most folks knew that he was going to win, but not by a landslide.
A pollster is a person who tells you what is going to happen before you know or they do. This forecaster has some tools and s/he works his or her magic. The Trump and Harris race was just the opposite of 1972 race. Trump won by 1% over Harris. Nathan Silvers of the “538” think tank ranks polls by their historical accuracy and how open they are to others and how they poll. There were 282 pollsters working on this race, and even more listed below them as they were not as prestigious Most hovered around 50/50. Some very prominent polls were wrong by about 1%-2%. Historians and journalists will tell us more precisely about it in the days ahead.
So, if you were the polling industry, you may want to know that you start with thinking that 95 times out of 100 times that you will probably be correct . There is a chart and it will be of assistance especially when you have millions of voters (Confidence Level.)
How much is the difference of votes between the 2 candidates? That was easy with the McGovern-Nixon race because there was a huge difference noted in earlier polls. So the final poll before the election was one where they would take a chance that Nixon would win between 5 times higher than McGovern or lower by 5 times. That is called Margin of Error. With Trump- Harris, we have gone through hell. It is 50/50. Accuracy is the question here.
The last is the sample size or should it be first? The bigger the sample size the more likely that we will lower the risk of being wrong. If we just ask a handful of people, we will increase the risk of being inaccurate. So in this current race, we talk to a lot of people. We are talking to about a thousand or more voters. Then folks vote. The pollsters then defend their findings.
Please recall that we want to be confident, accurate, and have a large enough sample size. This last election is very important because we may change the type of government and who may serve in it.
Joel Snell
319-270-6720
3105 Alleghany Dr. NE
Cedar Rapids, Iowa
52402-3315
joelsnell@hotmail.com