Model-based synthetic control methods

Presented by Gyuhyeong Goh, Kyungpook National University (South Korea)

Friday, August 23 2024
10:00 AM-11:00 AM ET
AUST 340
Webex Meeting Link
Coffee will be served at 3:00 pm in the Noether Lounge (AUST 326)

Synthetic Control Method (SCM) has increasingly been regarded as an innovative way to perform causal inference with a single treated unit. The SCM framework is commonly known as a model-free method since it constructs the so-called synthetic control unit based on the convex hull of the untreated units. Although several variations have been proposed to improve SCM, the field still lacks attempts to link SCM to statistical models. In this study, we develop a model-based SCM approach that enables us to not only improve the performance of SCM, but also perform statistical inference. In addition, we investigate theoretical properties of our model-based SCM. The performance of the proposed SCM is examined via comparative simulation studies and real data analysis. This is a joint work with Jae-kwang Kim (Iowa State University) and Jisang Yu (Kansas State University).

Speaker Bio:

Dr. Goh is currently an Associate Professor in the Department of Statistics at Kyungpook National University in South Korea. He received his PhD in 2015 from the Department of Statistics at the University of Connecticut under the supervision of Dr. Dipak Dey. During his PhD study, Goh received several awards, including Gottfried Noether Award (2012), NESS Student Research Award (2013), and JSM 2015 SBSS Student Paper Competition Winners. Soon after earning his PhD, Goh started his position at Kansas State University as an Assistant Professor of Statistics and promoted to an Associate Professor with tenure in 2021. His primary research interests are in developing Bayesian theory, methods, and computation for addressing high-dimensional data challenges and causal inference problems with observational data.