Sem (MA) Causal Identification
Dozent:innen: Dr. Holger ReinermannKurzname: Sem Causal Identific
Kurstyp: Seminar
Voraussetzungen / Organisatorisches
Wer in der ersten Sitzung unentschuldigt fehlt, verliert den Anspruch auf den Seminarplatz (Entschuldigung per Mail an Dozent*in ist ausreichend, kein ärztliches Attest erforderlich).Gültig für folgende vom Institut angebotenen Studiengänge:
- MA Politische Ökonomie Internationale Beziehungen (Modul 1 Politikwissenschaftliche Forschungsmethoden)
- MA Empirische Demokratieforschung (Modul 1 Politikwissenschaftliche Forschungsmethoden)
Inhalt
'Correlation is not causation' - we probably all have been confronted with this argument, and have learned to be wary of making causal claims based on mere associations between variables. Causal effects, we know, can only truly be identified through experiments.But it is also part of the truth that if political scientists only relied on experimental evidence, they might as well call it quits. Despite the impressive progress experimentalists have made in making the questions political scientists ask amenable to experiments, those remain infeasible, prohibitively costly, or unethical in many circumstances. So, often, we end up testing our hypotheses (typically derived from a theoretical argument that *does* invoke causality) with evidence that is ultimately observational in nature, or at least not truly experimental.
However, we do have some tricks up our sleeves to strengthen our empirical tests despite this problem. These range from specific data structures, such as panel data or natural experiments, to sophisticated estimation methods, such as difference-in-differences designs and instrumental variables. These and other approaches are the focus of our seminar. We will get to know different research designs, methods and frameworks that have been devised to enhance causal claims made from non-experimental data, discuss the assumptions they depend on and critically assess the value they add to empirical political science research. Crucially, we will also learn how to implement them in R/Stata (depending on the course attendees' preferences).
The seminar will be held in English - however, perfection is not required, only the ability to understand and make yourself understood. I will focus on application and include exercises with real data as much as possible. Therefore, you should have basic knowledge of one of the aforementioned software packages.
Zusätzliche Informationen
Hinweis für Soziologie Studierende: Abgabefrist für Hausarbeiten ist der 20.03.2024!Termine
| Datum (Wochentag) | Zeit | Ort |
|---|---|---|
| 24.10.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 31.10.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 07.11.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 14.11.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 21.11.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 28.11.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 05.12.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 12.12.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 19.12.2023 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 09.01.2024 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 16.01.2024 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 23.01.2024 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 30.01.2024 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |
| 06.02.2024 (Dienstag) | 14:15 - 15:45 | 02 601 Seminarraum |