by Professor Dr. Martin Spiess 

The course provides an introduction into linear models, their estimations and tests in the context of the analysis of panel or longitudinal data.

Panel data are characterized by observed units, e.g. individuals or firms, usually at equally spaced intervals over a certain time period.In contrast to models assuming independence of all observations, the repeated observations over time within the units are generally not independent in the panel data case.The richness of the class of panel models is the result of differing assumptions about the nature of the time dependence and the assumption about a possible dependence of the covariates included in the model and the error variables.

In addition to the lectures, accompanying exercises on the computer will illustrate the application of the methods. For these sessions the software STATA will be used.