Here is the CMT Uptime check phrase

A Conference on the ANES: Agenda

November 16, 2007

Gallup World Headquarters
Washington DC
United States

9:00-9:10 Introductions and Logistics
View the transcript  
Frank Newport, The Gallup Organization
Jon A. Krosnick, Stanford University
Matthew DeBell, Stanford University
David Howell, University of Michigan

Introduction to the American National Election Studies

History of the study, the study’s purpose, what data collections happened when, how data are distributed and when, how questionnaires are designed, how data are collected now, NSF funding support, CSES involvement, why we want advice from the news media pollsters.

Slideset: Catching up with the American National Election Studies 
Jon A. Krosnick, Stanford University
Arthur Lupia, University of Michigan


The British Election Studies and their relationship to the ANES

Harold Clarke, British Election Studies, University of Texas – Dallas
David Sanders, British Election Studies, University of Essex
John Curtice, University of Strathclyde
Nick Allum, University of Surrey


Examples of Lessons Learned from ANES Data Illuminating Voter Decision-Making

View the transcript

Sociotropic/retrospective voting, polarization over time in the American public, issue publics, the asymmetric nonlinear model of candidate evaluation, online vs. memory-based candidate evaluation, rationalization of party identification, Converse on ideology in the mass public, the structure of candidate personality trait evaluations, news media priming.

Slideset: Some of the Lessons Learned Using ANES Data 
Jon A. Krosnick, Stanford University
Arthur Lupia, University of Michigan


Long Pre- and Post-Election Face-To-Face Interviews

ANES conducts interviews lasting more than two hours with each respondent in a representative national sample (one hour plus before the election, and then another hour plus after the election with the same people). We have no trouble filling this many minutes of interviewing, tapping party identification, ideology, many evaluations of the incumbent administration, evaluations of the health of the nation in various domains, opinions on policy issues, perceptions of candidate stands on policy issues, the importance of each issue to each voter, overall attitudes toward the candidates and parties and social groups, voting intentions, election participation in other ways, demographics, and much more. To see examples of recent questionnaires, go to the 2000 Time Series Study page and click on the two links under the “QUESTIONNAIRES” heading on the left side. If you could collect 2 hours of info per respondent instead of just 15 minutes, what questions would you want to ask?

Slideset: Some of the Constructs Measured in the ANES 


Reinterviewing the Same People Monthly

If you were conducting a 19-month panel study, measuring the opinions of a single group of people once a month for 19 months, what ideas would you want to investigate? When would you like to make measurements? What would you measure how? Answering these questions of course entails addressing the viability of panel studies generally.


Self-administration and Video/audio Opportunities

If you could present visual displays, pictures, videos, graphs, etc. on a computer screen to the respondent or self-administer some questions, how would you use that opportunity?

Slideset: Examples of Questions Involving Visual Presentation and Self-Administration 

3:00-3:30 Questionnaire Design
View the transcript 


When measuring opinions, should we measure on continuous rating scales (as ANES has often done) or dichotomously (as the media polls often have)? On bipolar rating scales, should we offer midpoints or not? How should we choose the words to put in a question, when multiple synonyms could be used? What logic should we use to decide what order to ask questions in? How should we decide what policy issues to ask about? We will show many examples of ANES questions and media poll questions measuring the same things in different ways.

Slideset: Question Comparisons 


Questions on Terrorism

We will be asking a large battery of questions about issues related to terrorism. What should we ask about? We will inspire this discussion by presenting a large list of questions on terrorism topics from media polls.

3:50-4:20 Coordination of the ANES with News Media Surveys and Exit Polls
View the transcript  


Should we try to ask some questions that will be on your surveys done at the same time? If so, which questions? How should that be decided? What would the goal be of such coordination?


Can the ANES Collect Data to Help You Address Methodological Concerns in Your Own Studies?

For example, would it be of value to you for us to identify cellphone-only households in our 70%+ response rate 2008 face-to-face survey so you can compare those respondents to others in terms of political attitudes and behavior? Can we ask questions to assess whether respondents’ descriptions of what considerations shaped their votes (in your surveys) appear to validly match the results we get from elaborate statistical analyses using our data to identify the causes of vote choices?


How Should We Think About the Relation of the ANES to the Exit Polls?

Is there any reason to see a connection? Could the ANES help to serve any functions that the exit polls current serve for the news media?

5:20-5:30 Summary and Wrap-Up
View the transcript  


Frank Newport, The Gallup Organization
Jon A. Krosnick, Stanford University
Arthur Lupia, University of Michigan