A public opinion poll is an attempt to measure how a large group of people feel about specific political, social or economic issues. Polling organizations include private survey companies, universities and research centers. Many of these organizations conduct joint polls with news media, such as SSRS and CNN or NORC and the Associated Press or Siena College and The New York Times.
Many critics of polling believe that it is prone to biased results. They argue that the very act of asking questions can influence the answers, that respondents are not honest in their responses or that they may be influenced by “bandwagon” effects in which they see that their side is ahead in the polls and jump on the bandwagon. In the case of elections, some also believe that polls are designed to promote a particular point of view or candidate.
Even so, it is important to remember that polls are only one tool among many for learning about public opinion. Calls and letters to lawmakers provide a snapshot of opinions, but are limited to those with the time, energy and ability to craft and send them. Talking to friends or colleagues is another, but they are likely a self-selected sample that doesn’t represent the full range of experiences and viewpoints of the population at large.
In addition, the use of random sampling – giving every person in the population an equal chance of being included in the sample – is a method that can help to reduce bias in the survey’s results. Still, it is impossible to eliminate all bias. This is why the news media should always report a poll’s potential margin of error, as well as the name of the polling organization and the size of the sample.