|
Module Number:
| 14440
|
| Module Title: | Causal Data Science |
| |
Kausale Datenanalyse
|
| Department: |
Faculty 5 - Business, Law and Social Sciences
|
| Responsible Staff Member: | |
| Language of Teaching / Examination: | English |
| Duration: | 1 semester |
| Frequency of Offer: |
Every summer semester
|
| Credits: |
6
|
| Learning Outcome: | Students have a basic understanding of data science in the context of the identification of causal relationships. They are familiar with a verbal and graphical language to communicate about causality, and with key concepts, such as counterfactuals, outcome equivalence, and confounding effects. They know about typical classes of problems that do not allow causal interpretations of observed associations as well as typical solutions for these problems by means of data analytic and data collection methods. Moreover, students understand the tight interdependency of data analytics and the design of data collection to generate high-quality evidence and high-quality predictions. |
| Contents: | - Counterfactuals, Potential Outcomes, Causal Graphs, and typical problems (i.e., omitted relevant variables, measurement error, reverse causality, endogenous selection, endogenous treatment)
- Data analytic solutions: control variables, matching, weighting
- Data analytic solutions: instrumental variables, selection instruments
- Data collection solutions: real experiments
- Assumed experiments as mixed solutions: natural experiments, quasi-experiments, regression discontinuity
- Times series data as a mixed solution: diff-in-diff and related methods
- Reflections on moderation and mediation analyses, respectively, structural equation modeling
The module focuses on applications in business and economics, but the underlying theories and methods generalize beyond these fields. The course complements more traditional data science modules with a stronger focus on implementing data-scientific algorithms. Tutorials also apply these methods to the analysis of real-world problems with simulated and real datasets. Currently, the freely available software [R] is used in the practical parts of the tutorials. |
| Recommended Prerequisites: | - Basics of statistics, especially estimation and testing and simple regression analysis
|
| Mandatory Prerequisites: | None |
| Forms of Teaching and Proportion: | -
Lecture
/ 2 Hours per Week per Semester
-
Exercise
/ 2 Hours per Week per Semester
-
Self organised studies
/ 120 Hours
|
| Teaching Materials and Literature: | The lecture is based on selected chapters mostly from Morgan & Winship (2015). A few other articles or chapters will be provided during the module. Pearl, J. (2009) has become a classic reference in computer science. A more accessible introduction is found in Morgan & Winship (2015), the book on which most of the module is based. An accessible econometric perspective on some aspects of the module is offered by Angrist & Pischke (2014). More details on experiments can be found in Gerber & Green (2012).
- Pearl, J. (2009). Causality. Cambridge University Press
- Morgan, S. L., & Winship, C. (2015). Counterfactuals and causal inference. Methods and Principles for Social Research. Cambridge University Press.
- Angrist, J. D., & Pischke, J. S. (2014). Mastering'metrics: The path from cause to effect. Princeton university press.
- Gerber, A. S., & Green, D. P. (2012). Field experiments: Design, analysis, and interpretation. WW Norton.
A few additional shorter articles or chapters might be provided during the course of the module. |
| Module Examination: | Continuous Assessment (MCA) |
| Assessment Mode for Module Examination: | - 3 written partial examinations, 30 min each (each weighted 1/3)
|
| Evaluation of Module Examination: | Performance Verification – graded |
| Limited Number of Participants: | None |
| Part of the Study Programme: | -
Master (research-oriented) /
Angewandte Mathematik /
PO 2019
-
Master (research-oriented) /
Artificial Intelligence /
PO 2022
-
Master (research-oriented) /
Betriebswirtschaftslehre /
PO 2017
-
Master (research-oriented) /
Informatik /
PO 2008
-
Master (research-oriented) /
Transformation Studies /
PO 2024
-
Master (research-oriented) /
Wirtschaftsingenieurwesen /
PO 2019
|
| Remarks: | Tutorials are open to questions in English and German. |
| Module Components: | - Lecture Causal Data Science – 2 Hours per Week per Semester
- Exercice Causal Data Science – 2 Hours per Week per Semester
|
| Components to be offered in the Current Semester: | |
