A First-Time Course in Modern Causal Inference at the University of Costa Rica

• Activity Number: 85
• Type: Contributed
• Date/Time: Sunday, July 31, 2011 : 4:00 PM to 5:50 PM
• Sponsor: Section on Statistical Education
• Author(s): Eiliana Montero*+
• Companies: University of Costa Rica
• Keywords: Causal inference ; causal models ; Teaching ; Structural equation models ; SEM

Abstract

The experience of preparing and teaching a first time course in modern Causal Inference is described. The course is taught at a Master's program in Statistics at the University of Costa Rica. Besides regular students, continuing education students also attend. The instructor was recipient of the 2010-2011 Statistical Ambassador prize and preparing and teaching this course is one of her duties. Theorical issues regarding the nature and interpretation of causal evidence in non-experimental situations are discussed, and concrete applications using the Structural Equation Models (SEM) approach are presented, using LISREL and R software. Emphasis is put on how to conceive and operationalize relatively complex causal models in social science settings. Students are encouraged to develop and test a causal model using existing data as part of their final project in the course. Achievements and pitfalls of the course, from the instructor's point of view are also described, and recommendations for future experiences are given.