Putting Polls in their Place

By Mike Cronin

Now that election season is in full swing, we are subject to a constant stream of polls and polling results in the news. Many of these polls are legitimate attempts to gauge public opinion.  But some polls, and/or the results of some polls, are put to use to shape our opinions rather than discover them.  One example I heard recently alleged to be reporting opinions on who had little or no chance of winning the presidency in November.  The polling methods might have been legitimate, but the client who ordered the poll might have had an ulterior motive.  Here’s a possible scenario:

Candidate A, or an organization in favor of Candidate A, orders a legitimate poll from a reputable firm such as Gallup or Quinipiac to find out who people think have the worst chances of getting elected. The poll is taken and the results come back that people don’t think Candidates X, Y, or Z have much chance of getting elected in the fall.  Candidate A’s organization then trumpets these results loudly and often. Had the results been inconclusive, Candidate A’s organization wouldn’t have publicized the poll at all.

Knowing that we are all influenced to some degree or another by peer pressure, the results of the poll could sow discouragement among the supporters of Candidates X, Y, and Z, and even affect the candidates themselves, potentially hastening their departure from the race and reducing the competition against Candidate A.

There are dirtier tricks in the pantheon of polling practices. One such is to use the so-called “push-poll.” A push-poll is not really a poll at all in that it doesn’t seek to find out your opinion, but to change it or reinforce it. You might have received a piece of political junk mail that seems like a poll, but the questions will be highly “loaded” or biased.  Questions from such an artifact might read something like these examples:

“Candidate B wants seniors to lose their Medicare benefits. Do you think we should allow him to get his way?”

“The ABC project will pollute the local wetlands with 10x the current level of toxic substances.  Candidate B receives campaign donations from the company behind the project.  Don’t you think we need campaign finance reform?”

In both examples, the reader is first led to be outraged. The outrage is inherently tied to a candidate, then the respondent is manipulated to respond a certain way. The client of the poll wasn’t trying to find out people’s feelings on Medicare or campaign finance; rather, he or she was trying to sully the competition. A legitimate opinion poll will word the questions in as neutral a way as possible.  For instance, the questions above might be legitimized by rewording them in the following way:

“How do you feel about Medicare?

It should be maintained         Uncertain            It should be abolished”

“What is your opinion on campaign finance?

It needs to be completely reformed       It could use some reforms           It does not need any reforms”

So how can you peek behind the curtain of the polling game? Whenever you are asked to participate in a poll, or you hear polling results, consider these questions and/or look for trouble signals:

Who took the poll?  Was it a reputable polling firm?

Who commissioned the poll? What do they stand to gain or lose from the results?

Was the sample (the group of people taking the poll) drawn at random?  (If not, the poll will results will be biased, and therefore suspect.) Something to consider: most internet or call-in polls may seem to be sampling at random, but they are not. They are surveys of people who are already interested in the topic at hand! Most often, valid polls are accomplished via random digit dialing.

Was the sample size large enough?  (If the total population is, say, “likely voters,” which might be 75 million people, a sampling of ten people is not sufficient to gauge the opinion of the whole, but a sampling of 1000 or 10,000 might be. This is what statisticians get paid the big bucks to figure out.)

Are the questions biased or neutral?

Who is touting the results of the poll? Have they focused on the results of only one question that serves their ends, while ignoring the results of other questions on the same poll?

When you understand how polls can be used and misused, you increase your immunity to being manipulated in ignorance.

3 thoughts on “Putting Polls in their Place

  1. I found this interest and learnt a lot from reading this. I was wondering if you could check out my new politics blog and give it some support? I am just starting out and this blog is a way for me gain some writing experience.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s