9-30-08, 10:02 am
Editor's Note: This article is one of a series looking at the recent book The Opinion Makers, by David W. Moore and the validity of public opinion polls. See Riggins' earlier articles here and here.
In his book, The Opinion Makers, David W. Moore opens his third chapter by pointing out that large sections of the public know little, and care even less, about many of the issues that pollsters are asking them to give opinions about. Since the pollsters want to have dramatic splits in public opinion (it makes the poll more interesting for their media clients), they use questions with force-choice answers (i.e., usually two choices are given and 'unsure' is not given as a choice) and ignore the fact that many people don't know much about the issues. The polls thus often 'distort or completely mischaracterize what the American public is really thinking.' I almost said they 'misunderestimate' what is going on.
One of the tricks to get around a lack of public knowledge about the subject of the poll is to supply some information to the the person being polled. 'As you may know X has said that Y is the case. Do you agree with what X says or not.' But now you have biased the sample population you are polling by giving them this information. They no longer represent a typical cross section of the public. All polls do this and thus get 'a manufactured opinion based on a mythological public – measures that look like they represent what a rational, informed, and engaged citizenry might be thinking.'
The Gallup people tried to get more honest reflections of public opinion. George Gallup decided on a five question poll that would also measure what the public knew about an issue. Moore reproduces the results of a 1953 poll concerning support for the Taft-Hartley Act. The result was CHANGE IT 19 percent, LEAVE IT AS IT IS 11 percent, REPEAL IT 3 percent, NO OPINION 7 percent, NOT FOLLOWED THE ISSUE 60 percent.
This approach has not been adopted because the media clients of the polls don't consider it newsworthy to report that two-thirds of the public is not aware of the issues they are reporting on. Even though the polling companies know this type of poll is more accurate they have decided to rely mostly on the forced-choice method because it gets the big dramatic results their media clients want. This is shocking because they have no concept of the 'truth' but only want to sell their services to their media clients who also have no concept of 'truth' but only want more readers or viewers.
This chapter definitely points out the three main ways in which the polls falsify public opinion. 1) By not pointing out how much of the public is uninformed about the issues or doesn't care. 2) By using forced choice methods to get a response the pollster wants rather than what the person being polled would really have responded. This is a variant of the first way. 3) By supplying the person being polled with information he or she didn't have before (as a way of getting a 'choice') and thus biasing the sample. Most Americans would not have had the information that was supplied so the polling sample is not really representative. This 3rd point also could lead to a 4th in so far as the information supplied by pollsters is mostly an oversimplified presentation of the issues.
Moore ends his chapter with a quote from Daniel Yankelovich, a great pollster himself and with integrity: 'Sad to say, the media who sponsor opinion polls on policy issues have little or no stake in the quality of the poll findings they report.'
