Analysis of Panel Data: Practical Methods and Applications

The course focuses on econometric methods for analyzing panel data (longitudinal data) models. We will discuss important methods and illustrate these methods using real life data sets within Stata, and the popular open source software, R, enhancing student’s analytical and professional development within data-driven investigation.

Course level
Master, PhD candidates and professionals from all disciplines
Session 2
11 January to 18 January 2020
Co-ordinating lecturersAssoc Prof. Hande Karabiyik
Other Guest lecturersAssoc. Prof. Julia Schaumburg
Form(s) of instructionLectures, exercise sessions and computer classes
Form(s) of assessmentFinal report and team presentation
ECTS3 credits
Contact hours45 hours 
Tuition fee

€800 - non-VU students and staff

€500 - VU students and staff (includes VUmc)

We particularly encourage people with different backgrounds to apply so as to contribute to a lively and dynamic discussion. This course is designed for PhD candidates, master students, and professionals. In particular: 

  • PhD candidates from all disciplines, especially economics, finance, corporate finance, and business, who seek to develop an understanding of panel data models and the practical skills to analyze large data sets.
  • Master students who are aiming to pursue a PhD degree, or desire to work as a business practitioner with an interest in developing their ability to use econometric techniques while analyzing large data sets. 
  • Professionals who are in need of an in-depth understanding of panel data models, as well as the ability to guide themselves through projects that require the analysis of panel data sets.

Students of this course should have knowledge of basic econometric methods and basic linear algebra. In particular:

  •  Knowledge of linear multiple regression analysis; 
  • Knowledge of basic linear algebra; 
  • Basic knowledge of Stata, R, Matlab, or Python. 

If you are not sure whether your background is sufficient to follow this course, please contact us via h.karabiyik@vu.nl or graduatewinterschool@vu.nl. 

The course focuses on econometric methods for analyzing panel data (longitudinal data) models. We will discuss important methods and illustrate these methods using real life data sets within Stata, and the popular open source software, R, enhancing student’s analytical and professional development within data-driven investigation.

Students do not necessarily need to know R as a prerequisite – an introduction to using R for the analysis of panel data will be demonstrated and explained to you throughout the course. Students are free to choose any software to use while working on their assignment, including Stata, R, Python, and Matlab.

Being equipped with tools and knowledge to analyse panel data sets will advance many student’s professional careers tremendously, as it is a must-have knowledge in many fields of applied research. This will help students excel in the financial sector, in central banking, and in any field that is concerned with data analysis. 

A panel data set is a data set where there are observations of many units over multiple time points. For instance, the data set of the GDPs of EU countries over the last century is a panel data set. Using panel data sets is becoming more and more popular and those who can work with data are extremely sought-after, as the availability of such data sets are increasing every day within multiple fields of applied economic and financial research. 

The course will consist of three main parts. Each part will focus on a certain aspect of panel data analysis. Additionally, each part will contain computer exercise sessions, in which students will learn how to analyse panel data sets by using Stata and R functions, as well as commands that belong to current methods and applicable techniques. 

  • The first part of the course will focus on micro panels. Micro panels are panels where there are observations of many individual units, but only over a short period of time. For instance, household surveys are usually of this sort. They require specialized techniques to be analyzed. This part focuses on these specialized techniques and their applications. 
  • The second part of the course will focus on the diagnostic tests that analyze panel data sets. Panel data exhibits many features that may influence the ways required to estimate the models. We will discuss various tests that have been recently developed in the literature. This part will provide students with an in-depth knowledge of the theory of diagnostics of panel data sets and its applications.
  • The third part of the course will discuss the methods used to analyze large panel data sets. Large panels consist of many individual units and cover long time periods. These kind of panels require advanced time series techniques and other innovative techniques to analyze cross-sectional dimension. This part will mainly focus on methods that deal with cross-sectional dependence in various set-ups. Cross-sectional dependence occurs, for example, when the cross-section units are affected by the same unobserved shocks and is present in many panel data sets. There is a vast amount of literature on how to deal with the problems caused by cross-sectional dependence, and we will focus on these methods. This course will discuss how these methods can be applied practically, with illustrations, and with using real data.


By the end of this course, students will:

  • Have in-depth knowledge of how to analyze various types of panel data models;
  • Gain a thorough understanding of problems they may face while analyzing panel data sets;
  • Gain applicable knowledge about the solutions to the problems that can be presented while analyzing panel data sets;
  • Be able to choose appropriate tests while running diagnostics on panel data sets; 
  • Be able to choose appropriate methods to estimate panel data models;
  • Gain a recent and relevant knowledge of the current state of analyzing panel data models
  • Be able to propose potential empirical applications of panel data econometrics;
  • Be able to use Stata and R to analyze panel data sets;
  • Further develop professional programming skills. 

Course readings will be provided at the start of the course. The readings will consist of lecture slides provided by the lecturer, published research articles, and relevant book chapters.

Hande Karabiyik is an associate professor of Econometrics at Vrije Universiteit Amsterdam and a Tinbergen Institute fellow. She has obtained her PhD degree in Econometrics at Maastricht University in January, 2015. Upon obtaining her PhD, she worked as a post-doctoral researcher at Lund University, Sweden. She started working as an assistant professor at Vrije Universiteit, Amsterdam, in September, 2016. Her research focuses on the econometrics of cross-sectionally dependent panel data models. She published her work in leading journals of the field. She teaches courses within the bachelor programs of Econometrics and Operations Research and Econometrics and Data Science, as well as in the master program of Econometrics. 

Julia Schaumburg is an associate professor in Econometrics at Vrije Universiteit Amsterdam, a research professor at Halle Institute for Economic Research, a Tinbergen Institute research fellow, and a visiting researcher at De Nederlandsche Bank. She obtained a PhD in econometrics at Humboldt University, Berlin. Her research interests range from time series and financial econometrics, to systemic risk and financial stability applications. She obtained the prestigious Veni grant of NWO. Her work is published in various leading journals of the field. She teaches courses within the bachelor program of Econometrics and Operations Research.