By ANDY ANDREWS
Los Alamos World Futures Institute
The words of the title of this series all imply gathering and analyzing data of peoples’ opinions. Surveying might be about hard data such as age, height and weight, but it might be about opinions.
Do you prefer red or blue? What is your “view or judgment about something, not necessarily based on fact or knowledge?” (posted from Google). You do not have to provide evidence to support your opinion even though you might have it.
In polling a sample of a large population, supposedly random data, is gathered, but is affected by the process. In quantum mechanics it is called the Heisenberg Uncertainty Principle and classical physics it is the observer effect. Polling is neither quantum nor classical physics, but the opinion holder is affected by the measurer. The effect might be good or bad or even negligible, but there is an effect. Is the poll verbal, written, in person, via computer, or some other electronic device? Is it an intrusion or has the “pollee” volunteered? If one is attempting to determine a preference of a population from a sample, randomness is essential.
Statistics has a major role in polling but not in voting, maybe. In polling perhaps 1,000 people are asked for a response, assuming the population is significantly larger (330 million?). Per the mathematics of statistics for a binomial poll, this should lead to a result of plus or minus 3 percent. If the poll is preference of red or blue and the results show red at 48 percent and blue at 52 percent, in reality red falls in a range of 45 to 51 percent while the range for blue is 48 to 54 percent. Clearly there is room for error in interpretation. Plus, if the poll is merely 1,000 people, similar results can be expected only 95 percent of the time.
Today we see the results of polling almost every day if we pay attention. The reported results are usually in reference to a political race. While most polling organizations do so through “random” person to person contact, is it truly random? What do YOU say when you are contacted for your opinion? Do you utter some expletives and hang up or otherwise disconnect? Do you listen carefully and respond honestly? Or do you lie to the pollster because you are fed up? Or are you somewhere else? What is your view of the value of polling when the real results will be determined by voting?
Perhaps you decide to take the poll and you ask if you support candidate X because of his or her racist tendencies? There is no actual evidence of these tendencies and you hang up. As a result of the call, however, you may now begin to wonder about candidate X. This is an example of a push poll, an interactive marketing technique intended to manipulate or alter a potential voter’s beliefs – his or her opinion which does not demand fact or knowledge. Push polling is a form of spreading propaganda and rumors while appearing to be an opinion poll. But assuming it is recognized, can it affect valid polling?
We have been watching the polls leading up to election day and the actual voting. It has become almost similar to the NFL, NBS, NHL, and MLB. Voting day is the Super Bowl of the political race. OK, maybe just the local little league championship. Registered voters can cast a ballot if they choose.
Are those who choose not to vote the ones that that decline answering the poll? We do not know and it makes a difference because the election is the official and final poll, usually, generally, most of the time.
Unfortunately, elections are not always accurate if intended to determine the preference of the majority. If the voting is about two candidates or two proposals, what about the input of those choosing not to vote? We can say that it is too bad, they are ambivalent and the results are final. But what is the voting is about three or more candidate or propositions? Now it becomes problematic.
Assume there are three candidates or propositions: A, B, and C. If now of the candidates or propositions gets a clear majority, there can be a problem. Let us say that 100 votes are cast and A gets 46, B gets 44 and C gets 10. If the voters for C prefer A over B, A could get 56 votes. But if the voters for C preferred B over A, B would get 54 votes. You often hear arguments that we should have run offs all of the time, despite the fact that the process would be slowed down.
In the United States, as stated on Wikipedia, recording and reporting of presidential elections first occurred in 1824. Since then the winner of the election has often received only a plurality (less than 50 percent) of the vote but more than the other candidates individually. It has happened 19 times. Including 1824, by my counting, the United Sated has had 48 presidential elections and in half of them the winner had less than 50 percent of the vote.
That is 24 elections. In five of them the winner did not even have a plurality because the electoral college system affected the outcome, causing many vitriolic debates about states’ rights.
Obviously, the basis of voting and polling is complex. The math is confusing and difficult to understand even though we should fully comprehend statistics. Note the word “should.” It is easy to state an impossible level of attainment, it is more difficult to judge when necessary the statistical data presented and react appropriately. Surveying, choosing, and influencing is about us and what we are told or asked to do. We are statistics.
Till next time…
Los Alamos World Futures Institute website is LAWorldFutures.org. Feedback, volunteers and donations (501.c.3) are welcome. Email email@example.com or email firstname.lastname@example.org. Previously published columns can be found at www.ladailypost.com or www.laworldfutures.org.