METHODS FOR CAUSAL INFERENCE

Course objectives

Educational objectives The educational goal of the course is students' learning of the main statistical methods used for causal inference. That is, how to answer research questions about the impact of certain causes on a particular outcome. Knowledge and understanding At the end of the course, students know and understand the main methods for causal inference. Ability to apply knowledge and understanding Students learn how to apply the main methods for causal inference also through the use of a statistical software. Judgment independence The discussion of the various methods, even with team works, provides students with the skills necessary to analyze real situations critically and independently. Communicative skills Students acquire the basic elements for reasoning in quantitative terms about causal inference problems. These skills will be further developed through team works on real data. Learning skills Students who pass the exam are able to apply the methods learned in different application contexts.

Channel 1
ROBERTO ROCCI Lecturers' profile

Program - Frequency - Exams

Course program
Week Topic 1-2 Basic Definitions, Assumptions and Randomised Experiments 3-4 Non-Parametric Identification and Estimation 5 Selection on Observables: Matching, Regression and Propensity Score Estimators 6 Selection on Unobservables: Non-Parametric IV and Structural Equations 7 Difference-in-Differences Estimation: Selection on Observables and Unobservables 8 Regression Discontinuity Design
Prerequisites
No preliminary exams. However, it is necessary to know the fundamentals of calculus, probability calculus, descriptive and inferential statistics (general linear model included).
Books
- FRÖLICH M., SPERLICH S.(2019). Impact Evaluation.Treatment Effects and Causal Analysis. New York: Cambridge University Press. - Cunningham, S. (2021). Causal Inference: The Mixtape. Yale University Press. https://doi.org/10.2307/j.ctv1c29t27
Teaching mode
Theoretical lessons, discussions of exercises and applications on real data. Whenever possible, all the activities will be carried out in class.
Frequency
Students are strongly encouraged to attend the lessons.
Exam mode
The final exam is in two parts: written (80%) with exercises and theoretical questions, with open and close answer, oral (20%). The written part can be taken during the course with some intermediate tests.
Bibliography
- IMBENS G. W., RUBIN D. B. (2015). Causal inference for statistics, social, biomedical sciences: An introduction. New York: Cambridge University Press. - Scott Cunningham (2023). Causal Inference: The Mixtape. https://www.scunning.com/mixtape.html
Lesson mode
Theoretical lessons, discussions of exercises and applications on real data. Whenever possible, all the activities will be carried out in class.
  • Lesson code10612088
  • Academic year2024/2025
  • CourseStatistical Sciences
  • CurriculumData analytics
  • Year2nd year
  • Semester1st semester
  • SSDSECS-S/01
  • CFU9
  • Subject areaStatistico