I wonder how much credit people give these wildly inaccurate polls. Often the difference is less than the margin of error. They reach about 9% of the sample, according to The Week. Many people use only cell phones, and it is even harder to reach those people. This skews results even more. Although there is no credibility to these things, they may influence people. Is this right?
8 Answers | Add Yours
I would say that you are overstating the case against polls. It does not matter if the poll only reaches 9% of the sample. Polls are not meant to poll everyone but only a random sample. As long as the sample is random and has enough people in it, it can tell us something worth knowing. As for leads for one candidate or another that are within the margin of error, there are two responses. First, people should be well enough informed to be able to understand that a 1% lead for a candidate is within the margin of error. Second, if we look at the same poll over time, the trends should be accurate. In other words, it should be able to tell us if one candidate or the other is gaining.
If the polls are consistently close, as they are in this election, it should be obvious to everyone that it is a close election and the polls may not be able to predict who will win. But there is no real reason to ban polls. They give us important information about the attitudes of our fellow Americans and they help us to understand what is likely to happen in our political world.
The only problem I have with election polling is speaking with voters about how they voted. While there is a margin of error in any polling, I do not believe that that is the problem. For some, telling who they voted for is private. That said, these polls do give a heads up as to what may happen. I don't really see the harm in it.
As long as a poll operates from a good methodology, I think, as others have pointed out, that it is legitimate. Polling is a fairly advanced science these days, and while some polls are more accurate than others, it just demands a small amount of sophistication on the part of the voting public to make sense of them. It should also be noted that "banning polls" is probably unconstitutional under the First Amendment. I've attached an interesting article that grapples with the issue of polls actually affecting election results.
While at this point of the election season it can be easy to throw our hands up in the air and exclaim that polls should be outlawed, you haven't presented a strong enough argument. True, polls can be misleading, but isn't this true of all advertising? Polling should be taken as just a piece, albeit a small piece, of information when selecting a candidate. Sure, it'd be nice to say that no one voted for a candidate because of a skewed poll, but can we not also assume people vote for the candidate whose sign they saw in a neighbor's yard?
Voter education seems to be the larger issue. I would be interested in finding ways to ensure that the public knew about the candidates, but where would they go for fair and unbiased information?
I think the real issue at hand is the general public’s ability to interpret data. Unfortunately, most people have an extremely poor grasp of basic statistics and data collection practices. When someone doesn’t even understand something as basic as sample size, there is literally no way for them to independently evaluate the validity of any statistic that they are presented with.
My biggest issue with this type of polling is that the media uses it to generate a false sense of interest. It is in the best interests of the larger news organizations to manipulate the data in such as way as to make any particular race appear closer then it may actually be. If a news outlet reported that candidate A was beating candidate B by a healthy margin, then the public would lose interest and the media outlet would lose viewership.
where would our presidential elections be without the ever-present and always controversial polling? Polling is traditionally skewed and unreliable, but to argue that it should be outlawed altogther is a bit of a drastic stance. Ultimately, polling represents gathering and sharing of information, particularly that protected by the First Amendment, which in my opinion should not be limited in any particular way.
As mentioned in one of the above posts, polls can potentially influence constituents' mindsets, but even though that may be true, that unfortunate outcome is not grounds enough to endorse censorship or an enforced prohibition of polling.
I agree with #7. "Outlawing" polling would be a violation of the 1st Amendment, and add to the slippery slope of making free speech illegal. The price of everyone having a voice is that everyone has a voice, but the difference that many people don't understand is that the right to speak is not the right to be heard. If you don't like the polls, ignore them. They're not affecting you.
Having said that, I think that polls should be limited to actual scientific methods. Most polls are wildly inaccurate because of their methodology; even when they are proved to be largely correct in their predictions -- as those of the the most recent few months -- that does not mean they are accurate samples of the population. Better methods and more strict standards -- enforced by the polling centers, not by the government -- are necessary to create a useful set of data instead of just partisan echo-chamber.
While polling by phone and in person is annoying and often seems intrusive (not to mention inaccurate), the question for me would be why outlaw it? It is already illegal to do so within 50 feet (in most states) of an actual voting location, and there is certainly no requirement for people to answer polling questions, so I see no reason to legislate it out of existence.
Candidates and organizations use polling because they have a tendency to be accurate or at least in the ballpark in terms of the degree to which people support candidates or policies. If pollsters annoy you too much, you can always mislead them or just tell them you don't know who or what you support yet, as the less reliable polls become, the less likely they would be used as much.
We’ve answered 396,860 questions. We can answer yours, too.Ask a question