Political Polling Offers A Potentially Powerful Tool

In United States politics the importance of political polling in gauging voter opinion may often times be exaggerated. There are a number of things to consider in judging the effectiveness of a poll and it goes without saying that polls conducted on websites, whether personal or media-based, are not accurate samplings. (CNN, for instance, hosts a daily poll at the bottom of its home page seeking reader input on various subjects currently in the news spotlight.)

Any poll that is not based on random sampling will be skewed toward the audience of the hosting element. For instance, if a liberal blogger writes liberal political essays and offers a political polling application to his liberal readers it won’t be surprising that the poll results will be liberal leaning.

By the same token, political polling done by the candidates themselves will undoubtedly be produced with a campaign goal in mind – diverting attention away from a troubling topic or embarrassing incident, highlighting a campaign plank, or shoring up support in a key sector. When the candidate pays the market research company, it’s safe to assume the published political polling data will be worth the price of admission to the candidate.

Also consider how many people participated in the sampling. In a scientifically conducted poll where a professional pollster has carefully selected the participants to achieve a random sampling, the larger number of people responding, the better the results will be as an indicator of the opinion of the target group selected.

In most instances, in conducting the political polling, telephone interviews are used in exchanges that contain only residential accounts. While this goes a long way toward achieving the desired random sampling, no poll conducted in this way can ever be one hundred percent random when you factor in the people who live in the area but do not have phones, the fact that some residents will not be home at the time the poll is conducted, and the fact that some people are going to refuse to participate.

Any market research company in publishing a poll will give figures for margins of error. The accuracy of most polls is based on a confidence number or a sense of what the results would have been had it been possible to interview every single person in the target sampling of say one thousand individuals. Most polls operate in a confidence range of around ninety-five percent. With a plus or minus margin of error of three percent what the pollsters are saying is that if they had been able to interview that other five percent of potential respondents, their answers would have agreed or disagreed with the published results of the poll roughly three percent of the time.

In a nation as large and politically diverse as the United States, political polling is a tool with tremendous potential to gauge public opinion. No poll should be accepted at face value without exploring key points behind its administration. To truly evaluate the results, you must know more than the margin of error. You must know who paid for the poll and why, who conducted it, and how the respondents were chosen. Without a random sampling, there is no chance of having a usefully indicative poll.

Rate this article: Bad   Good

Send to friend Print version