Eilert Sundts hus
4th floor (map)
Moltke Moesvei 31
Lecturers: Professor Harry Ganzeboom, Department of Social Research Methodology,
Free University Amsterdam, & Dr. Paul Nieuwbeerta, NSCR, Leiden, The Netherlands
Main disciplines: Political Science, Sociology
Dates: 25 - 29 July 2005
Course Credits: 10 pts (ECTS)
Limitation: 20 participants
Requirements: Knowledge of multivariate regression in SPSS
Participants: Svetlana S. Babenko, Johannes Bergh, Kirill Kalinin, Rune Karlsen, A. Jan Kutylowski, Daniel Scheuregger, Tim Spier. Lecturers: Professor Harry G.B. Ganzeboom and Dr. Paul Nieuwbeerta. Including: Summer School organizer Karl Olav Grønlund Sørensen.
Political research has shown consistent and convincing evidence on the relationship between class and status indicators and voting behaviour. Questions on the relationship do not concentrate on whether and why, but more on how strong and when. Over the last ten years in particular there has been a vehement debate on the decline of class voting: whether it has occurred, where and when the decline would have been stronger of weaker / absent, and how class voting depends on the measurement of class. All of these questions address comparative issues, there are questions about historical changes and/or differences between countries/systems of different political orientation.
The present course looks at the comparative analysis of the class/vote connection from a methodological angle. In particular we discuss (A) the various ways of conceptualising and measuring social class, (B) the comparative measurement of political preferences, and (C) the use of powerful XT (pooled time series or repeated cross-section models), to help to solve the various issues at stake. Students are expected to be empirical researchers and will learn how to do large scale comparative research hands-on. They should be proficient in multivariate regression in SPSS or a similar statistical package (the course practices will use SPSS as the tool of analysis) and preferably have knowledge of multinomial and binomial logistic regression models.
The course will consist of the following 10 modules
Topic 1: Class and Vote - setting a stage. This lecture discusses the historical background of the problem and highlights the literature of the past 10 years.
Topic 2: Class and Vote - the empirical problem. This lecture scrutinized the recent empirical contributions to the discussion and advances testable hypotheses for the comparative analysis of the class-vote relationship.
Topic 3: Principles of comparative measurement. The comparative measurement of social class: class typologies versus continuous scales. The use of ISCO as a harmonization tool for comparing occupational titles. EGP, ISEI, SIOPS.
Topic 4: Introduction to the International Stratification and Politics File. Computer exercises on the standardization of class measures.
Topic 5: The comparative measurement of political preferences: scaling parties. Assessing party positions from external indicators and by using scales association models. Conditional multinomial logit models.
Topic 6: Computer exercises including ordinal and condition multinomial logit models. Contrast with the linear regression model.
Topic 7: Problems and principles of large scale comparative analyses. The degrees of freedom problem. Statistical power. Trends models using synthetic cohorts. Combining cross-sectional and time-series comparisons in XT models.
Topic 8: Computer exercises with XT modelling.
Topic 9: A comparison of comparisons: how do results depend upon (A) class measures used, countries included, (C) scaling of party preferences, (D) model of comparative analysis?
Topic 10: Assignments: student presentations of historical developments in their own country. Assignments for the take-home essay.
Basic literature (provisional)