Advanced statistical methods in social sciences 1600-SZD-ID-ZMS
The course covers topics in advanced regression analysis. It begins with linear regression, including all of its assumptions and a thorough understanding of the potential problems that a researcher may encounter when using this method (e.g., when analyzing interaction effects). Next, nonlinear models for nominal variables (logistic regression), count data (Poisson and negative binomial regression), and the McFadden choice model are discussed. Finally, quasi-experimental approaches are covered (e.g., regression with fixed effects).
Course coordinators
Learning outcomes
Knowledge | The graduate knows and understands:
WG_02 - the main development trends in the disciplines of the social sciences in which the education is provided
Skills | The graduate is able to:
UW_01 – make use of knowledge from various fields of science, in particular the social sciences in order to creatively identify, formulate and innovatively solve complex problems or perform tasks of a research nature, and in particular to: define the purpose and object of scientific research in the field of the social sciences, formulate a research hypothesis; develop research methods, techniques and tools and apply them creatively; make inferences based on scientific findings
UK_04 - participating in scientific discourse in the field of the social sciences
Social competences | The graduate is ready to
KK_01 - critically evaluating achievements within a given scientific discipline in the field of the social sciences
And others: 1. Knows all 10 assumptions of the linear model. 2. Is able to interpret the effects obtained using linear regression. 3. Can analyze and interpret interaction effects in linear regression. 4. Knows the limitations of using the linear model to analyze binary variables. 5. Can interpret the effects obtained using logistic regression. 6. Knows the differences in interpreting interaction effects between linear and nonlinear models. 7. Knows the limitations of using the linear model to analyze variables that are natural numbers. 8. Can interpret the effects obtained using Poisson and negative binomial regression. 9. Understands the problem of overdispersion and underdispersion in the Poisson model. 10. Can analyze choice problems using the McFadden model. 11. Knows what the independence of irrelevant alternatives assumption is. 12. Can estimate and interpret a linear model with fixed effects.
Assessment criteria
Description of requirements related to participation in classes, including the permitted number of explained absences: At least 80% participation (16 hours)
Principles for passing the classes and the subject (including resit session): Participation in class and final empirical research project
Methods for the verification of learning outcomes: Final empirical research project
Evaluation criteria: At least 80% class participation and satisfactory quality of the empirical research project
Bibliography
1. King, Gary. "How not to lie with statistics: Avoiding common mistakes in quantitative political science." American Journal of Political Science (1986): 666-687. 2. Long, J. S. (2014). Regression models for nominal and ordinal outcomes. The SAGE handbook of regression analysis and causal inference. 3. Rabe-Hesketh, S., & Skrondal, A. (2008). Multilevel and longitudinal modeling using Stata. STATA press. 4. McFadden, D. (1972). Conditional logit analysis of qualitative choice behavior. 5. Górecki, M. A., & Kukołowicz, P. (2018). Electoral formula, legal threshold and the number of parties: a natural experiment. Party Politics, 24(6), 617-628.
Additional information
Additional information (registration calendar, class conductors, localization and schedules of classes), might be available in the USOSweb system: