Advanced regression techniques

Third-cycle level | 7.5 credits | Course code: SF30002
VT 2021
Study period: 2021-01-25 - 2021-03-22
LANGUAGE OF INSTRUCTION: The course is given in English
Application period: 2020-11-16 - 2020-12-07

Course description

The course focuses on the application, diagnosis and interpretation of advanced regressiontechniques for social sciences data. Particular emphasis is placed on data processing, modelspecification, diagnostics and inferencing.


Requirements and Selection

Entry requirements

Completed courses in simple regression analysis and introduction to multiple regression analysisat the equivalent of second-cycle level.

Qualifying applicants are persons admitted to postgraduate studies at the University ofGothenburg or another university.


Place on the course is primarily given to persons admitted to postgraduate education at the Faculty of Social Sciences, then, in turn, persons admitted to the University of Gothenburg, Chalmers University of Technology and other universities in or outside the country.

Other information

The course consists of five components that are examined separately:

Statistical analysis, data processing and model specification (1 credit)

This course component begins with an introduction to syntax-based regression analysis followedby data processing and the handling of ‘missing’ data. This course component concludes with anintroduction to model specification focusing on causal analysis and statistical inferencing asregards measurement errors, model specification, and omitted variable bias.

Multiple regression analysis and logistic regression analysis (1 credit)

This course component includes linear regression analysis (OLS regression) with moreexplanatory variables (independent variables) and the handling of dichotomous dependentvariables (e.g. great/small confidence in politicians). The specification of models, regressiondiagnostics, interaction effects and assumptions are also treated.

Regression discontinuity design, instrumentation and two-step regression (1.5 credits)

This course component includes regression discontinuity design, which is a regression-basedquasi-experimental research design aimed at isolating causal effects. This course component alsofocuses on the use and implementation of instrument variables and two-stage OLS regression.

Multi-level analysis (2 credits)

This course component includes an introduction to multi-level analysis, i.e., when data is available at multiple levels (e.g. municipality, unit, individual). The specification of basicmodels, estimation of parameters, model comparisons and assumptions are also treated.

Panel data analysis (2 credits)

This course component includes techniques for analysing data that has been observed over time(i.e. to study changes and dynamic processes). The specification of models, correction of errors,causality and assumptions are also treated.



Course coordinator Nicholas Charron,

Course administrator Anne-Marie Deresiewicz,

Course syllabus


Reading and reference list

Reading and reference list for the course


Department of Political Science


Social Science

Type of course

Method course



CONTACTNicholas Charron