Skip to main content icon/video/no-internet

Exit polls are in-person surveys in which data are gathered immediately after people have engaged in the behavior about which they are being surveyed, such as voting in an election. The survey methods that are used in exit polls apply to the measurement of a wide variety of behaviors, but in the minds of most people exit polls are most closely associated with what is done on Election Day to help project the winning candidates before the final vote tally is announced. Although political exit polling is done in many countries, it is the exit polling conducted for elections in the United States that is covered here.

How Exit Polling is Conducted and Used in U.S. Elections

The exit polls that are conducted nationwide and in most individual states for the general election in the United States are among the largest single-day surveys that are conducted anywhere, with data from more than 100,000 respondents being gathered, processed, and analyzed within one 24-hour period.

To estimate the outcome of an election in a particular geopolitical area of the United States, which most typically is done at the state level, a stratified random sample of voting precincts within the area is selected, and at least one interviewer is sent to each of the sampled precincts. In the 2004 U.S. general election, there were 1,469 sampled precincts nationwide, and in 2006 there were 993. Those exit polls were conducted by Edition Media Research and Mitofsky International, the organizations that were hired to gather the exit poll data for their news media funders (ABC, the Associated Press [AP], CBS, CNN, Fox, and NBC). On a systematic basis, and in order to obtain a completed questionnaire, the exit poll interviewer stops (i.e. intercepts) people who just finished voting as they exit from their voting places. For example, the interviewers may do this with every 10th person who comes out of the voting place. In each sampled precinct, an average of approximately 100 voters is interviewed over the course of Election Day. Not all exiting voters who are stopped agree to complete the exit poll questionnaire, but in those cases the interviewer records basic demographic information about these refusing voters. This information is used later as part of analyses that investigate the nature of exit poll nonresponse. Interviewers at each sampled precinct telephone in the data they are gathering at three scheduled times on Election Day: mid-morning, early afternoon, and within the hour before voting ends in the precinct.

In order to gather the exit poll data, the interviewer typically hands the selected voter a questionnaire on a clipboard and asks her or him to complete it and then deposit it in a survey “ballot box.” The questionnaire gathers three types of data: (1) it measures who the sampled citizen voted for in the day's key election contests (e.g. president, senator, and governor); (2) it measures various attitudes held by the voter that the news media sponsors believe will be associated with the votes cast by the sampled voter; and (3) it measures key demographic and lifestyle characteristics of the voter to further help explain why he or she voted as he or she did. All of these data are gathered via a questionnaire comprised of 20–25 questions that fit onto one piece of paper, which is printed on the front and back sides.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading