ABOUT THE DATASET:
Time Series Study
Completions: 1,807 pre-election; 1,555 post-election
Sample: all fresh cross-section
Modes used: face-to-face, telephone
Weights: V000002, V000002a
STUDY CONTENT HIGHLIGHTS:
In addition to content on electoral participation, voting behavior, and public opinion, the ANES 2000 contains questions in other areas such as values and predispositions, media exposure, social altruism, and social networks. Special-interest and topical content includes a sizeable battery on the Clinton legacy and a smaller retrospective battery on former President George Bush Sr., new social trust questions specific to neighborhood and workplace, expanded content on civic engagement, questions related to the debate about campaign finance reform, and the first ANES time series appearance of measures on cognitive style.
STUDY DESIGN HIGHLIGHTS:
The ANES 2000 design included a mode experiment utilizing a dual frame sample. Some respondents were selected by traditional area probability sampling and were interviewed face-to-face. Other respondents were interviewed over the telephone, having been selected by the first-ever use of RDD sampling in an ANES time series study. Pre-election interviews averaging 68 minutes were conducted between September 5 and November 6. No interviewing was conducted on election day, November 7. Post-election interviews averaging 64 minutes were administered November 8 through December 18. Half-samples were used at times to reduce overall interview length and to facilitate administration of mode-effect experiments. The mode experiments sought to examine response order effects and different question designs for branching versus scale formats, among other experiments. Randomized assignment to half-samples, random question order within batteries, and randomized assignment to alternative battery and question placements were made possible by CAI instrumentation, which differed by mode as necessary for administration of the questions. Specific design features included a new measure of voter turnout addressing the issue of over-reporting. Additionally, placement of vote report was moved to earlier in the interview to eliminate possible contamination from other questions related to candidates. Last, ballot cards were redesigned; respondents were randomly assigned to one of two ballot cards presenting, by office, either Democratic or Republican candidates listed first.
The total sample included 2,984 eligible persons and produced 1,807 pre-election interviews. In the post-election survey, 200 randomly selected persons who provided a face-to-face pre-election interview were assigned to be re-interviewed by telephone - of those, 168 granted a post-election interview. Altogether, 1,555 persons provided post-election interviews.
V000002 is a full-sample pre-election weight which is the product of the household non-response adjustment factor, the within-household selection weight (number of eligible persons), and a post-stratification adjustment factor by age and education. V000002a is the corresponding post-election weight, adjusted for attrition.