The Gap-Closing Estimand
March 9th from 2:30 -4 pm
Target Audience: This workshop is designed for researchers familiar with basic statistics (e.g., regression). Programming examples will use R, but the underlying causal concepts will be accessible to those unfamiliar with R.
Description: Disparities across race, gender, and class are important targets of descriptive research. But rather than only describing disparities, research would ideally inform interventions to close those gaps. This workshop will introduce causal decomposition analyses as developed in biostatistics and in the social sciences, the causal assumptions those analyses require, and methods available for estimation.
– Conceptualize disparities in a causal framework
– Formalize causal assumptions in Directed Acyclic Graphs
– Use the gapclosing package to conduct analysis
– Visualize results
Speaker: Ian Lundberg, PhD, is an Assistant Professor of Information Science at Cornell University, working in an interdisciplinary department home to computer scientists and social scientists. Ian is a sociologist by training whose research develops statistical methods and applies those methods to questions about inequality, poverty, and mobility.