Study Planning

The planning and specification of an Election Study is a collaborative effort involving the Principal Investigators, the NES Board of Overseers, and the research community. Planning takes place over a full two years prior to conducting the field work, and at each step there are multiple opportunities for the research community to help specify the content of the survey. The planning process for the 1994 National Election Study, described below, exemplifies this process, the broad outlines of which have been in place since 1977.

Planning for the 1994 National Election Study began in November 1992 when the Board sent a stimulus letter to the NES mailing list of 2300 social scientists inviting suggestions on new themes and questions to be piloted in 1993 with an eye towards possible inclusion on the 1994 study. At its February 1993 meeting, the Board reviewed the responses and selected four social scientists to join with a subset of the Board and the Principal Investigators to plan the 1993 NES Pilot Study. The Pilot Study Planning Committee met in May of 1993 and later continued discussions through an email conference, drafting a questionnaire for review by the full Board of Overseers at its June 1993 meeting.

The Telephone Facility of the Survey Research Center conducted the Pilot Study in the fall of 1993, with interviewing ending November 28th. Responsibility for analyzing and reporting on Pilot Study data rests with the individuals whose research and development projects have been included on the survey. The Board of Overseers reviewed these reports to appoint a 1994 Study Planning Committee and formulate recommendations to that Committee for the content and design of the 1994 National Election Study.

Content for the 1994 study was drawn from two sources: new instrumentation developed in the 1993 pilot work, and "core" Time-Series questions. The new instrumentation has undergone extensive testing on the previous year's Pilot Study. Alternative question wordings have been tried, the reliability and validity of each item established. The NES Board, which bears responsibility for deciding which core Time-Series items will be carried in each National Election Study, reviewed the core items and proposed which ones should be continued and/or dropped. These recommendations were circulated to the NES research community to solicit their reactions and suggestions. At their June 1994 Board meeting, the Board revised the list of core items in response to the suggestions it received.

The 1994 Planning Committee designed the 1994 National Election Study. This Board-appointed committee was comprised of the Principal Investigators, a subset of the NES Board and several members of the NES community. The Planning committee met in late April 1994, to review the Pilot Study reports and other recommendations from the community and to draft the 1994 questionnaire. The NES Staff worked closely with the Committee to prepare, distribute, and time the draft questionnaire, as well as to monitor the budgetary implications of the Committee's deliberations. The full Board of Overseers then reviewed the Planning Committee's recommendations.

During the late summer of 1994, the draft instrument was pretested, by administering the questionnaire to about 30 people. The pretests were recorded. The Staff and field section personnel debriefed the interviewers to understand their experiences with the instrument, and assess the implications of the pretest timings for the length of the production interview. The pretest was also "behavior coded" -- i.e., specially trained coders listened to the recorded interviews and noted each instance where interviewers departed from question text, or where respondents had difficulty understanding the question or gave an unresponsive answer that was not noted by the interviewer.

Results of the debriefing and the behavior coding informed the deliberations of the Board of Overseers at its September meeting. There, final decisions about study content were made, including cuts needed to accommodate the budgeted interview length. The instrument itself continued to be proofed extensively until mid-September, when it went to press. The Board created a small committee to make final decisions about special additions to the questionnaire needed to accommodate important domestic or international political changes. Such changes can ususally be implemented at the last minute, or even as the study is in the field.

The 1994 Election Study: Planning Activities