Gesis Spring Seminar 2024

26 Feb - 01 March, 2024

Lecturers:

Alexandru Cernat

Nick Shryane

Short Course Description:

Longitudinal data are an essential tool for researchers as they can help answer questions about change in time, causal relations, and the timing of events. They come in many shapes, from traditional panel surveys to social media and sensor data. Because of their additional complexity, specialized statistical models are needed to analyze them.

In this course, you will learn how to analyze longitudinal data using R. The course is developed to include statistical models from a number of different fields, giving students a comprehensive knowledge of models such as: multilevel models for change, latent growth models, cross-lagged models, and survival models. The course is also hands-on, each topic being accompanied by real world applications using R and practical exercises. In addition to learning about statistical models, the students will also learn how to prepare and visualize longitudinal data. They will also have the opportunity to discuss their own research projects and get guidance on how they can use the methods covered in the course in their own work.

04 - 08 March, 2024

Lecturers:

Scott Cunningham

Teaching Assistant:

Kyle Butts

Short Course Description:

When researchers are not able to field randomized experiments to study the causal effects of large social programs
due to their size, associated costs, feasibility, and ethical constraints, they often rely on natural experiments such as
law changes or natural disasters. The most popular research design for estimating the causal effects using
longitudinal data is the difference-in-differences design. The method is extremely popular in the empirical social
sciences.  For instance, around 25% of all papers at the NBER working paper series using difference-in-differences.  

But, while difference-in-differences is relatively straightforward, unbiasedness in the parameter estimates depends
on the setup of the quasi-experiment and the methodology used. For instance, when there are more than one dates
when units are treated, then traditional panel methods are no longer guaranteed to be unbiased, even under parallel
trends. Our understanding of these issues has evolved considerably over the last several years, both in terms of
econometric theory and software implementation. This workshop will review this emerging work covering both the
intuition behind the statistical models and the technical details of the models themselves using lectures, discussion
and group exercises using R and/or Stata.

11-15 March, 2024

Lecturers:

Martin Spindler

Jannis Kück

Short Course Description:

Participants of this course will learn and apply recent Causal Machine Learning methods to analyze effects in either cross-sectional or panel data. Causal Machine Learning combines the field machine learning, which was developed for predictions and is based on correlation, and the field of causal inference. In this course we will focus on the so-called Double Machine Learning approach (DML) which allows for valid inference in high-dimensional settings.

This course will focus on tools that are easy to implement for practitioners in the R / Python and covers three blocks:

1. Basics of causal inference
2. Basics of machine learning
3. Double Machine Learning (DML) for cross-sectional and panel data (including difference-in-differences, instrumental variables and mediation analysis)