Public opinion polls are used daily to gather information about the attitudes of a population regarding politics and other social issues. These polls supposedly represent how the general public feels as a whole about particular topics. While they often prove to be useful in determining outcomes of elections or persuading politicians or business owners to take a particular action on an issue, the public opinion poll is a flawed process that has its own unique set of disadvantages.
The Leading Option
According to Surveys.com.au, one major disadvantage of public opinion polling is the tendency of the person taking the survey to go with "the leading option." The leading option is the answer to a polling question that the researcher suggests is the popular answer while asking someone else. For example, a pollster may tell a person that research has shown that Candidate A is the most common answer when people are asked who they think will win the election between Candidates A, B and C. When the poll question is actually asked, it is phrased, "Among Candidate A, B and C, who do you believe will win the election?" Having already heard what the supposed most popular answer is, those who do not have a strong opinion about the matter are likely to go with Candidate A because most other people have apparently done the same. This leads to inaccurate public opinion and is a way that pollsters with an agenda can help get the results they want.
If you've seen poll results on the news, there is usually a disclaimer that lets the viewer know that there is a plus or minus 3 per cent margin of error. This may occasionally be the case, but there is no solid way to tell just how big the error margin is, according to PollingReport.com. Sampling errors take place in a number of different ways. If a pollster is conducting a sidewalk survey, there is a strong possibility several people will refuse to take part in it. If the poll was about attitudes toward public opinion polls, for example, a very significant portion of the population may not be represented. And, of course, there is outright dishonesty by pollsters. If the poll is driven by an agenda, there is very little to stop them from doctoring the results to fit that agenda or wording questions in a way that is likely to provoke a certain response.
Selection bias happens when the people intentionally selected to take part in a poll may not be representative of the entire population. If a conservative talk radio station is conducting a call-in opinion poll from its listeners concerning their opinion on a liberal political candidate, the results are fairly predictable. The liberal politician is likely to be viewed as unfavourable to the station's conservative listeners. In addition, in the above example, the people who feel strongly about a subject are likely to call in multiple times to vote. According to The Skeptic's Dictionary website, a poll conducted by Alfred Kinsey about homosexuality showed that 10 per cent of the American population is gay. Later studies suggested the number is more like 2 per cent. Kinsey's numbers were subject to selection bias because he conducted the survey with people in prisons and those who attended his lectures. Neither was representative of the entire population.