How the Nevada exit poll was conducted

Wednesday, Nov. 8, 2006 | 7:20 a.m.

The poll of Nevada voters was conducted for the National Election Pool - The Associated Press, ABC, CBS, CNN, Fox and NBC - by Edison Media Research and Mitofsky International.

The survey included an exit poll of Election Day voters and telephone interviews with those who voted early or absentee.

The exit poll was conducted Tuesday at a randomly selected sample of 20 precincts around Nevada. A statewide telephone poll was conducted during the past week to interview absentee voters. The absentee interviews were weighted to represent 42 percent of the overall sample, Edison/Mitofsky's estimate of the absentee or early vote in the state electorate.

In the exit poll, as people left the polling places, Edison/Mitofsky interviewers asked them to fill out a confidential paper questionnaire prepared by NEP representatives. The interviewers selected voters at a set interval - such as every fifth person - so that each participant had an equal chance of being picked. Voters interviewed by phone were asked the same questions.

The exit poll results were adjusted to reflect the different probabilities of selecting a sample precinct and people attending each, as well as by the observed sex, race and estimated age of voters who refused to participate. The telephone survey results also were weighted to geographic region, household size and the number of telephones in the household.

As with any survey, the results could vary because of chance variations in the sample. For this poll of 1,702 respondents - including 703 absentee voters interviewed by phone - there was one chance in 20 that sampling error would cause the results to vary by more than plus or minus 4 percentage points from the opinions of all Nevada voters.

Sampling error also depends on how many exit poll sites have voters with the characteristic of interest. For example, black or high-income voters may be found clustered in only a few sample precincts. Sampling error may be up to three times larger for clustered characteristics.

Polls are subject to other sources of error, such as from question wording or order.