ABOUT THE DATASET:
Type of study: ANES Time Series Study
Sample universe: U.S. eligible voters
Sample composition: panel (2000 Time Series) and fresh cross-section cases
Sample selection: RDD (Random Digit Dialing)
Number of waves: 2 (pre-election and post-election)
Completions: 1,511 pre-election; 1,346 post-election
Modes used: telephone
Instrument format: CATI (Computer Assisted Telephone Interviewing)
Weights: V020101 , V020102
Additional files: contextual file
Associated studies or derivative files: ANES 2000-2004 Merged File, ANES 2004 Panel Study
STUDY CONTENT HIGHLIGHTS:
In addition to content on electoral participation, voting behavior, and public opinion, the ANES 2002 contains questions in other areas such as social trust and civic engagement. Special-interest and topical content includes questions on the terrorist attacks of September 11, 2001, the war on terrorism, economic inequality, the 2000 Presidential election, recent corporate scandals, the 2001 tax cut, and proposed elimination of the estate tax.
STUDY DESIGN HIGHLIGHTS:
The ANES 2002 is the first midterm time series study to include a pre-election interview in addition to the standard post-election interview, and is the first ANES time series study to be conducted entirely by telephone. Interviews using the 36 minute pre-election instrument were conducted September 18 through November 4. No interviewing occurred on election day, November 5. The 32 minute post-election interview was administered November 6 through December 6. Half-samples and quarter-samples were utilized to reduce overall interview length, with random assignment of alternate question and word orders in many sections. A special experiment addressed the issue of over-report of turnout, with random assignment of either the experimental format of the turnout question that was introduced in the ANES 2000, or the standard format of the turnout question. Data collection was conducted by the Indiana University Center for Survey Research (www.indiana.edu/~csr) using CATI instrumentation.
The sample consisted of 1,807 respondents who had previously provided an interview in the ANES 2000 Time Series Study ('panel' respondents), together with fresh RDD cross-section cases, including 921 eligible persons, for an eligible total of 2,733. The total sample of panel and fresh cases in this study constitute an unweighted cross-section. Of these,1,511 completed pre-election interviews (1,187 panel, 324 fresh cross-section) and 1,346 were re-interviewed in the post-election study (1,070 panel, 276 fresh cross-section).
V020101 is a full-sample pre-election weight which compensates for differential sampling or inclusion probabilities for respondents and for differential nonresponse across geographic sample design categories. It also provides a post-stratification adjustment by age and education groups. V020102 is the corresponding post-election weight adjusted for attrition. These weights may also be used by analysts interested in analyzing only the data for the panel respondents re-interviewed in 2002.
The 2002 Auxiliary File of contextual data is a file primarily intended to provide assistance to analysts gathering contextual information. Included are biographical variables for candidates and incumbents, as well as variables describing incumbents by political record and ranking. This file is downloadable from the ANES Data Center.
ASSOCIATED STUDIES OR DERIVATIVE FILES:
The ANES 2004 Panel Study conducted re-interviews of 'panel' respondents from the ANES 2002 Time Series Study. The ANES 2000-2004 Merged File creates a 2000-2002-2004 panel file that includes all interviews from the ANES 2000 Time Series Study and reinterview data for 'panel' respondents from the ANES 2002 Time Series Study, together with data from the ANES 2004 Panel Study. These two associated datasets are downloadable from the ANES Data Center.