Government » American Politics of Public Opinion » How Is Public Opinion Measured?

Summary and Main Ideas


The purpose of a poll is to identify how a population feels about an issue or candidate. Many polling companies and news outlets use statisticians and social scientists to design accurate and scientific polls and to reduce errors. A scientific poll will try to create a representative and random sample to ensure the responses are similar to what the actual population of an area believes. Scientific polls also have lower margins of error, which means they better predict what the overall public or population thinks. Most polls are administered through phones, online, or via social media. Even in scientific polls, issues like timing, social pressure, lack of knowledge, and human nature can create results that do not match true public opinion. Polls can also be used as campaign devices to try to change a voter’s mind on an issue or candidate.

Practice Questions

  1. Why do pollsters interview random people throughout the country when trying to project which candidate will win a presidential election?
  2. How have changes in technology made polling more difficult?

Sample Answer:

1. If a pollster interviews only a certain type of person, the sample will be biased and the poll will be inaccurate.


Bradley effect: the difference between a poll result and an election result in which voters gave a socially desirable poll response rather than a true response that might be perceived as racist

exit poll: an election poll taken by interviewing voters as they leave a polling place

leading question: a question worded to lead a respondent to give a desired answer

margin of error: a number that states how far the poll results may be from the actual preferences of the total population of citizens

push poll: politically biased campaign information presented as a poll in order to change minds

random sample: a limited number of people from the overall population selected in such a way that each has an equal chance of being chosen

representative sample: a group of respondents demographically similar to the population of interest

straw poll: an informal and unofficial election poll conducted with a non-random population

[Attributions and Licenses]

This is a lesson from the tutorial, American Politics of Public Opinion and you are encouraged to log in or register, so that you can track your progress.

Log In

Share Thoughts