|To:||The ANES Research Community|
|From:||Mark Hansen, Chair, ANES Board of Overseers|
Arthur Lupia, Chair, 2004 ANES Planning Committee
Nancy Burns, Principal Investigator
Donald Kinder, Co-Principal Investigator
|Re:||The 2004 National Election Study: Process and Planning|
|Date:||December 11, 2003|
Dear NES User Community,
The Board of Overseers of the American National Election Studies is in the midst of planning the 2004 Election Study, the latest in a continuous series of presidential election studies now spanning more than fifty years.
We write to inform you about current plans for the 2004 Study and to seek your input. We will also tell you about a new website whose sole purpose is to enhance communication between planners and potential users of the 2004 study.
At present, the 2004 ANES survey will consist of 1200 cases with 100 minutes of interview time per respondent. All interviews will be conducted face-to-face using the same probability area sampling design as in the past. As usual, the pre-election interviewing will begin early in September and continue until the day before the election. Pre-election interviews are 50 minutes long. Post-election interviewing will begin the day after the election, and we expect approximately two-thirds of the interviews to be completed over the following three weeks. Post-election interviews are 50 minutes long.Study Content
Some of the instrumentation for the 2004 study will be chosen with the upcoming election in mind. Such instrumentation may or may not have been relevant in previous elections or have appeared in previous studies. As always, however, a sizable fraction of the 2004 survey will consist of "core": concepts and questions that have appeared regularly over the years and that provide a basis for systematic analysis of political continuity and change on an ever-lengthening time scale. Core is not absolutely fixed, but the Board considers additions to it only under special circumstances - which include new theoretical or conceptual developments, methodological innovations, or changes in the political world.
A challenge for ANES planners in 2004 is that current funding levels mandate shorter interviews than before. The 100 minutes of interview time per respondent is more than 30 minutes shorter than the face-to-face interview time in 2000. To increase the number of cases and the interview length, Nancy Burns and Don Kinder (the ANES principal investigators) are pursuing additional funding from a variety of sources. One of these pursuits would allow re-interviews of the panel respondents from 2000 and 2002 – an exciting possibility for scholars who want to gauge the impact of important political and foreign policy events since those times. Similar efforts by Burns and Kinder funded the entire 2002 National Election Study. Their success in coming months will determine whether the 2004 study has more cases or longer interviews.
At present, however, the Board of Overseers and the Planning Committee will proceed on the basis of existing funding. Therefore, it proposes to reconcile the difference in interview length by cutting the interview time devoted to core items from 70 minutes per respondent to less than 60 minutes per respondent.
Please note that if Burns and Kinder are not able to secure additional funding, then any increases to the proposed 2004 core must come at the expense of new questions specifically designed to capture important attributes of the 2004 elections or of other "non-core" questions that have rotated on and off past studies.Core Review
At its November meeting, the Board completed its review of core with these circumstances in mind. It identified questions that should be included at the discretion of the Planning Committee (a.k.a. inventory) rather than mandated to appear on NES surveys as part of core. It also proposed soliciting views from the user community about how another set of questions can be asked more efficiently or effectively.
The number in parentheses following each of the questions below refers to the identification number on the NES Core Spreadsheet (see the new interactive version at [ADDRESS DELETED; NO LONGER IN USE]). The spreadsheet lists all items included in core and specifies in which previous ANES study each has appeared. The spreadsheet also lists questions and concepts that the Board has previously designated for "rotation": that is, to be asked occasionally, for time-series purposes, but not routinely.
The questions proposed for change in status from "core" to inventory are:
See note about starred items below.
In light of changing patterns in political communication and media usage, the Board recommends that the question "Did you see any information about this election campaign on the Web?" (7014, asked since 1996) be promoted from inventory to core.
The board also recommends consolidating the questions:
The Board also requests information from the user community about potential substitutes for the following questions:
In recent years, question 3000's response variance has dwindled and remained small. Low variance, in turn, reduces the potential inferential power of these questions. We ask the user community to propose more effective ways of addressing matters of gender politics and policy and to provide arguments about the importance of such items relative to other topical areas that the ANES can pursue.
The Board also recommends changes to the following sets of questions:
The Board also asks the user community to comment on the effectiveness and efficiency of the ANES questions about
On this topic, it recommends that three of the following six questions be cut from core:
The Board seeks information about whether each of these items provides independent value to a broad community of users and about the extent to which scales made from all of these questions outperforms the best three-item scales. Specific examples will be helpful. More generally, the Board seeks input from users about which of the religion questions (which also includes questions 9017-9024) should be given the highest priority and retained in core.
In the event that question 4100 is retained, the Board asks whether there is a more effective way to ask the "Born-Again Christian" question. We have received conflicting reports about whether the term has been eclipsed in common parlance by terms such as "Evangelical" and whether the information can be derived from questions 9022-9024.
The Board also seeks advice on the value of the group closeness questions (8011). Have they been successful in providing to scholars informational value above and beyond that conveyed by the long-standing feeling thermometer questions (8000-8010)? Specific examples will be helpful.The ANES and the Comparative Study of Electoral Systems
The American National Election Studies is part of an international consortium called the Comparative Study of Electoral Systems (CSES). All member nations commit to run a common set of items to enable and enhance comparative election research. Our plan is to nest the new CSES module within the 2004 study – to this end, Burns and Kinder, Russ Dalton, and Phil Shively have submitted two grant proposals to add minutes to the 2004 study for this purpose. To learn more about CSES, visit http://www.cses.org or view the module at http://www.cses.org/resources/module2/m2micro.txt (the 2004 ANES will contain only those questions pertaining to the U.S. electoral system).Note About Starred Items
The CSES module contains slight variations of the starred questions listed above. So while the Board recommends that these questions be removed from core, it also recommends that such questions be asked in 2004 as part of NES's commitment to CSES.Tell Us What You Think
We are soliciting your advice concerning these Board recommendations. Tell us what you think about core, and any other advice you might have about the 2004 Election Study. There are several ways in which you can participate in this process.
The best way to offer advice is through an email address dedicated to the 2004 study. It is [ADDRESS DELETED; NO LONGER IN USE]. Additional opportunities for correspondence will appear in mid-December, with the launch of a new website, [ADDRESS DELETED; NO LONGER IN USE]. This site is dedicated to providing information about the 2004 study and to making it easier for the user community to offer constructive advice. It will include message boards, interactive utilities about the design of the 2004 study, and a question usage poll. This site will remain open until the Fall of 2004.
You can continue to consult existing ANES resources at the ANES website, http://www.electionstudies.org. In addition to containing data from previous ANES surveys, this site also houses the newly-created "Questions Asked in ANES Surveys" which contains a comprehensive inventory of questions asked over the years http://www.electionstudies.org/resources/questions/questions.htm. To view recent questionnaires, visit
If you have trouble downloading these documents or the core spreadsheet from the Web, please contact the NES2004 Project Staff at [ADDRESS DELETED; NO LONGER IN USE].
The next major step in the planning process will be the meeting on February 6-8 of the 2004 NES Planning Committee, chaired by Arthur Lupia. In addition to Board members, the Planning Committee consists of David Brady of Stanford University, Raymond Duch of the University of Houston, Kathleen McGraw of the Ohio State University, Robert Shapiro of Columbia University, and Daron Shaw of the University of Texas. The committee will make a detailed recommendation about the content of the 2004 study, which the Board of Overseers will review at its meeting in May.
Given the reduction in available interview time, the Planning Committee must recommend cutting some non-core questions that appeared on earlier surveys. Note also that while the Board's recommended cuts in core increases the Planning Committee's discretion, they do not prevent the Planning Committee from including such questions in its proposal for 2004. The Planning Committee, therefore, will appreciate not only appeals to include certain questions in 2004 but also arguments for excluding particular questions. In all cases, such appeals are more informative when accompanied by concrete demonstrations of the questions’ importance to a broad scientific community.
Important decisions are ahead of us. We want to hear from you.