This course focuses on statistical methods for causal inference, with an emphasis on how to frame a causal (rather than associative) research question and design a study to address that question. What implicit assumptions underlie claims of discrimination? Why do we believe that smoking causes lung cancer? We will cover both randomized experiments – the history of randomization, principles for experimental design, and the non-parametric foundations of randomization-based inference – and methods for drawing causal conclusions from non-randomized studies, such as propensity score matching. Students will develop the expertise necessary to assess the credibility of causal claims and master the conceptual and computational tools needed to design and analyze studies that lead to causal inferences. Examples will come from economics, psychology, sociology, political science, medicine, and beyond. Previous exposure to the statistical software R is expected; students who have not previously coded in R may enroll but should expect to put in additional effort to learn this skill.

Units: 1

Max Enrollment: 15

Crosslisted Courses:

Prerequisites: Any one of QR 260/STAT 260, STAT 318, ECON 203, SOC 290, PSYC 305 or a Psychology 300-level R course; or a Quantitative Analysis Institute Certificate; or permission of the instructor.

Instructor: Pattanayak

Distribution Requirements: SBA - Social and Behavioral Analysis

Degree Requirements: DL - Data Literacy (Formerly QRF); DL - Data Literacy (Formerly QRDL)

Typical Periods Offered: Every other year

Semesters Offered this Academic Year: Spring

Notes: