Course Description
Design and analysis of social experiments; review of diverse examples of experiments from around the world; reasoning about cause and effect.
Athena Title
Social Experimentation
Pre or Corequisite
INTL 3200 or INTL 3200E or INTL 3300
Grading System
A - F (Traditional)
Course Objectives
When this course is finished, students should: - Understand the distinction between observational, experimental, and quasi-experimental studies - Understand the fundamental barriers to inference from observational data - Understand how experimental studies overcome this barrier - Gain familiarity with the potential outcomes framework and the Rubin Causal model - Gain experience in the design of experiments, including practical knowledge of randomization, factorial designs, delivery of treatment, etc. - Gain basic fluency with the statistical methods used to analyze experiments - Gain basic fluency with statistical software used to analyze experiments
Topical Outline
Foundations - What is causal inference and why is it important? - What sorts of questions can we ask about the social world? - Which questions have causal answers? Framework - Why is it difficult to answer causal questions with observational data? - How do experimental studies solve the fundamental obstacles to inference encountered by observational studies? - Introduction to the potential outcomes framework - Introduction to randomization and experimental design Keystone examples of experimental design - Tennessee STAR experiment - RAND Health experiment Review of experimental designs in key areas - Experimental designs to understand health - Experimental designs to understand discrimination - Issues in the implementation of experimental design, including non-compliance and interference
Syllabus