Learning about cause and effect from data can be tricky--especially when studying humans. Two of this year's Nobel laureates in economics, Josh Angrist and Guido Imbens, devised a host of statistical techniques for disentangling causation and correlation in social and biomedical sciences. This talk reviewed Angrist and Imben's most highly-cited paper, the classic "Identification of Causal Effects Using Instrumental Variables" (1996, written with statistician Don Rubin). Examples of instrumental variables in the wild--some clever and convincing, others (in my opinion) rather dubious were provided, as well as an introduction of principal stratification, a generalization of Angrist, Imbens, and Rubin's groundbreaking work.
Adam Sales is an assistant professor of statistics in the WPI Department of Mathematical Sciences, and the Learning Sciences and Technologies and Data Science programs. His research focuses on statistical methods for causal inference using large, messy datasets, mostly from applications in education science. He will be teaching a course in the spring semester on causal inference and survey sampling (MA 547) and does MQPs and IQPs studying and applying statistical models.