Issues in Analyzing Data from the Dual-Mode 2000 American National Election Study

For scholars interested in methodological issues concerning survey response and the analysis of incomplete survey data, the 2000 American National Election Study (ANES) is a gold mine of research opportunities. However, unless researchers proceed with caution, the riches of the 2000 ANES may turn into fool’s gold. The 2000 study differs from previous ANES studies by inter-viewing respondents in three different ways: individuals were contacted and interviewed over the telephone, in person, or a combination of the two. In addition, the questionnaire contained numer-ous experimental manipulations. The combination of the dual-mode design and the experimental manipulations creates several potential pitfalls for the researcher using the 2000 ANES that he or she may not typically encounter. Before proceeding to discuss the issues surrounding the 2000 ANES, it may be useful to offer the following suggestions with regards to the dual-mode design. If the researcher seeks to maintain continuity and consistency with research that has utilized studies from previous years, the best solution is to use only the face-to-face respondents. If the researcher seeks to analyze the impact of congressional district level variables on individual behavior and attitudes, it may be better to analyze the respondents that were interviewed over the telephone in the pre and post-wave of th