Select indoor events are open to the public, including some performances, lectures and athletic competitions. Visitors without a Williams ID must show proof of vaccination for entry and masking is required for all. Guests are also welcome in outdoor spaces.
See the Visit page of our Covid website for full details.
Examining the Structure of Bias Using Directed Acyclic Graphs by Prof. Rohit Bhattacharya
Examining the Structure of Bias Using Directed Acyclic Graphs by Prof. Rohit Bhattacharya, Williams College, Statistics Colloquium, Wednesday, October 13, 1:10 – 1:50 pm, North Science Building 017, Wachenheim
Abstract: The study of qualitative and quantitative theories of causation is a fundamental scientific endeavor. However, data from the real world present several challenges to drawing valid causal inference — these include confounding bias, selection bias, and censoring bias (including the systematic censoring of data in administrative records.) In this talk I will discuss the use of causal directed acyclic graphs (causal DAGs) as a succinct visual representation of the structure of these kinds of biases. I will then discuss how causal DAGs can be used to augment the statistician’s toolkit to establish confidence in (or evaluate) the causal validity of data analyses.