A stochastic model of preference change and its application to 1992 presidential election panel data

Bibliographic Data

ID7528670
AuthorsMichel Regenwetter (0000-0002-5607-9016, Duke University, corresponding author), Jean-Claude Falmagne, Jean‐Claude Falmagne (University of California, Irvine), B Grofman (0000-0002-2801-3351, University of California, Irvine)
Year1999
Volume106
Issue2
Pages362-384
Publication date1999-04-01
Open AccessNo
TypeARTICLE
VenuePsychological Review (JOURNAL)
Journal identifiersISSN: 0033-295X • E-ISSN: 1939-1471
PublisherAmerican Psychological Association (PUBLISHER • US)
DOI10.1037/0033-295x.106.2.362
OpenAlexW1978478987
LanguageEN
Citations received8
References cited13

The authors present and test a model for the evolution of preferences. Personal preferences are represented by rankings with possible ties and are posited to change under the influence of of information in the environment. These tokens may not be directly controlled or observed by the researcher. The authors apply the model to 1992 National Election Study panel data (W. E. Miller, D. R. Kinder, S. J. Rosenstone, & NES, 1993). The parameter estimates suggest that negative campaigning played a major role in the information flow. Democrats and Republicans experienced a barrage of contradicting information about Perot; Democrats, Republicans, and Independents each received or perceived different information. A shift in the perception of the candidates led the Republicans to evaluate Bush and Perot less favorably after the election. These results demonstrate the model's potential to analyze persuasion as a real-time stochastic process and without a media content analysis

Econometrics · Economics · Microeconomics · Panel data · Political science · Politics · Preference · Presidential election · Presidential system · Random effects model · Computer Science · Electoral Systems and Political Participation · Medicine · Opinion Dynamics and Social Influence · Psychology · Social Media and Politics

Tools

Open DOISci-Hub