Simon Hug
Regression, missing data and more
Simon Hug is lecturer at the Department of Political Science at the
University of Geneva, Switzerland and currently visiting scholar at the
University of California, San Diego. He holds a PhD. from the University
of Michigan and taught at the University of Michigan and at the University
of Geneva. His research interests include the formation of new political
parties, the effect of institutions, and more particularly referendums and
federalism, on decision-making and conflict resolution, formal theory and
quantitative methods. His publications include articles on party
formation, green politics, referendums and research methodology in the
"European Journal of Political Research," "Public Choice," "Mobilization,"
"Comparative Political Studies," the "Journal of Conflict Resolution,"
"Party Politics," and other journals as well as in several edited volumes.
He is co-author with Stefano Bartolini and Daniele Caramani of "Political
Parties and Party Systems. A
Bibliographic Guide to the Literature on Parties and Party Systems in
Europe since 1945" (London Sage, 1998) and the author of "Altering Party
Systems. Strategic Behavior and the Emergence of New Political Parties in
Western Democracies" (Ann Arbor University of Michigan Press,
forthcoming)
The workshop aims at providing the student with a solid knowledge of two
important topics in statistical analysis, namely the classical linear
regression model and the problem of missing data. Consequently, the
starting point is a thorough review of the classical linear regression
model, its underlying theory and its crucial assumptions. Violations of
these basic assumptions will be discussed in detail, and a special
emphasis will be given to problems of missing data. In order to
understand the problem of missing data and the possible solutions, some
basic knowledge of the extensions of the linear model and the basic
nonlinear models (e.g., probit and logit) is necessary. Thus, the
workshop also provides an overview of these different models with
limited or qualitative dependent variables. While the theoretical
material will be covered in class, lab sessions will force the students
to work in a hands-on fashion on the various topics. Several exercises
cover the different topics and special emphasis will be given to the
interpretation of statistical results.
Consequently, the main objective of the workshop is to impart a thorough
knowledge of a classical tool of empirical analysis, namely the linear
regression model. At the end of the workshop, students should have a
firm grasp of the conditions under which this tool performs well and
when it fails. Both in theory and in practice they should be able to
identify possible problems and adopt the appropriate remedies.
Similarly, problems of missing data should no longer be an obstacle in
the students' own empirical work. They should be able to assess the
possible biases introduced by various ways in which one can deal with
missing data and be able to adopt a strategy that solves their problems.
Bibliography
Basic text/overview
- Achen, Christopher H. 1986. Statistical Analysis of
Quasi-Experiments. Berkeley University of California Press.
- Greene, William H. 1990. Econometric Analysis. New
York MacMillan Publishing Company, ch.20-21.
- Gujarati, Damodar N. 1998. Essentials of Econometrics
2nd edition. New York McGraw-Hill, ch.5-14.
- Gujarati, Damodar N. 1995. Basic Econometrics. 3rd
edition. New York McGraw-Hill, ch.9, 15, 16.
- Hanushek, Eric A.; Jackson, John E. 1977. Statistical
Methods for Social Scientists. New York Academic Press.
Remedial Reading
- Achen, Christopher H. 1982. Interpreting and Using
Regression. Beverly Hills Sage Publications.
- Gujarati, Damodar N. 1998. Essentials of Econometrics.
2nd edition. New York McGraw-Hill, ch.1-4.
- Lewis-Beck, Michael S. 1980. Applied Regression.
Beverly Hills Sage Publications.
Prerequisites
Some notions of probability theory, statistical inference, basic calculus
and matrix algebra (Achen, Gujarati and Lewis-Beck cover most of these
things, while Appendix II in Hanushek and Jackson is a concise review of
matrix algebra)