Abstract
This paper presents a Thurstonian model for the analysis of preference change. Preferences are expressed in the form of rankings, possibly with ties. A vector-autoregression framework is used to investigate relationships between past and current rankings. It is shown that this approach yields a parsimonious and easily interpretable representation of individual preference differences in time-dependent ranking data. A detailed analysis of the 1992 National Election Study illustrates the proposed approach.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 300-314 |
| Number of pages | 15 |
| Journal | Journal of Mathematical Psychology |
| Volume | 46 |
| Issue number | 3 |
| DOIs | |
| State | Published - 2002 |
Funding
This research was partially supported by NSF Grant SBR97-30197. I thank the Action Editor, William H. Batchelder, for helpful comments. Adress correspondence and reprint requests to Ulf Böckenholt, Department of Psychology, University of Illinois, Champaign, IL 61820. E-mail: [email protected].
Keywords
- Maximum likelihood estimation
- Parameter-driven process
- Ranking data
- Ties
ASJC Scopus subject areas
- General Psychology
- Applied Mathematics