1988-1992 Merged Senate File
About the Dataset
- Type of Study: ANES Merged File
- Sample universe: eligible voters in U.S. households with telephones
- Sample composition: target number of IWs by state (50 states)
- Sample selection: per component study
- Number of waves: 1 each study
- Completions: per component study
- Modes used: telephone
- Instrument format: CATI
- Weights: PS0002
- Auxiliary files: none
Study Content Highlights
This release is an ANES derivative file which merges data from 3 studies: the ANES 1988 Senate Study, the ANES 1990 Senate Study, and the ANES 1992 Senate Study. The 1988, 1990 and 1992 Senate Election studies were very similar in content and design, each producing survey data which facilitate the comparison of House and Senate races through the use of questions that generally parallel the questions used in Time Series studies since 1978 concerning respondents’ interaction with and evaluation of candidates for the United States House of Representatives. The data also allow comparison of perceptions and evaluations of Senators in states where there was no Senate election alongside evaluations and perceptions of Senators running for re-election.
Topics covered included: respondent’s recall and likes/dislikes of House and Senate candidates, issues discussed in the campaigns, contact with House and Senate candidates/incumbents, respondent’s opinion of the proper roles for Senators and Representatives, a limited set of issue questions, liberal/conservative self-placement, party identification, media exposure, and demographic information as well as other areas. All variables from the component studies are included.
Study Design Highlights
Each of the 1988, 1990, and 1992 Senate studies comprised a single wave of interviewing and utilized an independent sample. The average lengths of the three Senate surveys were 32 minutes, 41 minutes and 34 minutes, respectively. In 1988, the target number of interviews for each state was 75. In 1990, the target number of interviews in states with Senate elections was 75; in the 15 states without Senate elections, the target number of interviews was 48. In 1992, the target number of interviews was 56 interviews per state whether or not there was a Senate race in the state.
Each study employed a different algorithm for selecting household telephone numbers. The 1988 Senate study utilized a Waksberg-Mitofsky design, in the first stage of which working blocks of household numbers (hundred-number blocks) were sampled with probability proportional to size (PPS). Among these primary sampling units, 25 were chosen for each state, and then a selection of telephone numbers from each HNB was chosen by a restricted simple random sampling procedure. The 1990 Senate Election Study used, within states, a “PPS-to-listed counts” two-stage design, by which approximately half of the sample was selected from a frame of listed numbers and half was generated using an RDD procedure from the set of hundred series in the list sample. In the 1992 Senate Election Study, a “list assisted” or “listed hundred series” RDD sample design was employed for one-stage selection of telephone numbers without clustering, selecting telephone numbers using the GENESYS RDD sampling system through Marketing Systems Group. 1992 samples of telephone numbers were selected from a frame consisting of all the possible numbers that could be generated from hundred series that had at least two listed household phone numbers.
The nine questionnaire versions used in 1988 and 1990 corresponded to different combinations for types of Senate and House races. In 1992, with reduced House-level question content in comparison to the 1988 and 1990 Senate studies, only 2 questionnaires were necessary: one for California, which had two Senate races, and one for all other states. Interviews were conducted November 14-December 20, 1988 for the 1988 Senate study; for the 1990 and 1992 Senate studies, interviewing was conducted during the time periods of November 7-December 22, 1990 and November 4-December 8, 1992, respectively. There were 3,145 completed interviews in the 1988 Senate study, 3,349 in the 1990 Senate study, and 2,759 in the 1992 Senate Study.
The relative sample weight variable (VPS0002) should be used for all descriptive analyses comparing two or more states or one state at two points in time. The relative weight values are based on the original RDD and dual-frame probabilities and do not incorporate any adjustments for differential non-response or telephone subscription rates across the states; the weighted totals for states may not be in proportion to the state household populations.
Auxiliary File Description
No auxiliary file is present for this derivative file however contextual data are merged with the release file, including election returns, voting indices, and information about the Senate.