3403768: Multiple lineare regression

back to overview
Semester:WS 12/13
Scheduled in semester:1
Semester Hours per Week / Contact Hours:40.0 L / 30.0 h
Self-directed study time:60.0 h

Module coordination/Lecturers


Master's degree programme in IT and Business Process Management (01.10.2008)


  • Simple linear regression
- Least squares methods
- Parametric rating, t-tests, confidence intervals, p-values
- ANOVA table
- R^2 coefficient
- Residual analysis
  • Twofold linear regression
- Parametric rating, t-tests, confidence intervals, p-values
- ANOVA table
- R^2 coefficient
- Comparison between simple and twofold regression, distortion through omission of variables
- quadratic regression
  • multiple linear regression
- parametric rating, t-tests
- confidence intervals, p-values
- ANOVA table
- R^2 and corrected R^2 coefficient
- Dummy variables
- Gauss-Markov theorem
- Multicollinearity
  • Model selection
- Transformations
- Selection of variables

Lecture Goals

The lecture "multiple linear regression" focuses on both basic concepts and procedures of one of the most ultilised statistical techniques in practical situations. The acquired competences for the individual learning contents are as follows:

Students are able to carry out linear regressions by hand bya applying the methods of least squares. In keeping with test statistics and confidence intervals, they are able to both calculate and control the model preconditions by making use of the residues.

Using the twofold linear regression as an example, students are able to point out distortions that can emerge through omitted variables. They are aware of in how far multiple linear regression is actually an extension to a simple linear regression.

Using the Gauss-Markov Theorem, students are able to ground the wide spread of the least squares method. By making use of dummy variables, they are able to integrate categorical explaining variables into the model. Furthermore, they are acquainted with the problematicity of multicollinearity.

With respect to method selection, students employ the common transformations and interpret the results correctly. Moreover, they make use of the common strategies in order to build models by way of variable selection. Students solve relevant example from the practice by using a statistical software. By doing so, they are able to interpret the respective outcomes correctly and are aware of the constraints that preside over the interpreted results.


Lectures Method

Interaktive Vorlesung



  • Wooldrige, J.M. (2006): Introductory Econometrics, 3. Aufl., Thomson, Kap. 1-7.

  • Montgomery, D.C.; Peck, E.A.; Vining, C. G. (2006): Introduc-tion to Linear Regression Analysis, 4. Aufl., Wiley & Sons.
  • Maddala, G.D. (2001): Introduction to Econometrics, 3. Aufl., Wiley & Sons.


Den Studierenden werden zusammenfassende Folien und/oder ein Kurzskript sowie Übungsaufgaben zur Verfügung gestellt.


02.11.201209:00 - 12:00H2
09.11.201209:00 - 12:00H2
16.11.201209:00 - 12:00H5 (Fabrikweg)
07.12.201209:00 - 12:00H5 (Fabrikweg)
13.12.201213:30 - 16:45S10 (Fabrikweg)
14.12.201209:00 - 12:00H5 (Fabrikweg)
21.12.201209:00 - 12:00S4
21.12.201213:30 - 16:45S4
12.01.201309:00 - 16:45H2
18.01.201310:00 - 12:00H3