Preface

Welcome to the research module in econometrics & statistics! Here is my contact information:

Two R exercise sessions will be held by one of my PhD students:

The pdf version of these lecture notes is on the class website. The html version is better formatted and can be found here.


Description: This class will introduce you to research in econometrics and statistics and will help you prepare for a master thesis in that area. The course has several parts.

Theory: We will discuss fundamental concepts of econometrics and statistics, including estimation theory and inference in a general setting as well as the linear regression model as a specific example. We will review large and small sample properties of estimators, confidence intervals, and hypothesis testing. Many of these results will carry over to settings beyond the linear regression model. We might also briefly introduce modern statistical methods, including nonparametric and high-dimensional methods.

Implementation: Using the statistical programming language and software environment R, we will illustrate the theoretical results using simulated and real data. The focus is on verifying theoretical results and investigating finite sample properties of statistical methods using Monte Carlo simulations. These simulations are typically an integral part of a master thesis in econometrics and statistics (and often something students struggle with).

Projects: For the projects, you will have the opportunity to choose among a set of specific projects. Additionally, you can suggest your own topics, but we will have to assess their feasibility. Each project focuses on one specific statistical method/topic (for instance quantile regression, bootstrap, etc.) and must contain the following two parts:

  1. A precise description of the statistical method and its theoretical properties.
  2. A Monte Carlo simulation study to assess the finite sample properties.

It is also desirable, but not required, to provide an exemplary real-data application to showcase the practical use of the statistical method. If you prefer, you can instead do a deeper simulation study. These simulations can be as complicated as you wish because there are always additional aspects one could consider.

For the projects you can work in groups of up to three students (see next section for more details).


Software: Due to the mathematical contents, it is strongly recommended to use LaTeX for preparing the presentation slides and the term papers. Moreover, the Monte Carlo simulations and the real-data applications will make it necessary to work with advanced software such as R, Python, or Matlab. We will give a short introduction to R, but we generally expect that you are willing and able to learn R independently. We will also provide a very brief introduction to LaTeX.


Important: You need to register for this course via BASIS. Exam registration for this class is much earlier than for other classes.


Time Table: Below is tentative overview of the topics that we plan to cover during each lecture.

Date Topic Instructor
Oct 8 General Introduction / Estimation Theory Freyberger
Oct 9 Estimation Theory Freyberger
Oct 15 Monte Carlo Simulations Kappenberg
Oct 16 Monte Carlo Simulations Kappenberg
Oct 22 Hypothesis Testing Freyberger
Oct 23 Hypothesis Testing / Regression Freyberger
Oct 29 Regression Freyberger
Oct 30 Advanced Topics Freyberger
Jan 21 Presentations
Jan 22 Presentations


Supervision meetings: In November, December, and January you can make appointments to discuss your projects during the regular lecture hours here.