In May 2020, CSBS hosted a methods series workshop called Causal Inference with Observational Data. The workshop kicked off with an introduction to causal inference by CSBS Director Brent Roberts.
Video playback of the workshop is now available on Media Space, segmented accordingly:
Jonathan Livengood, Associate Professor of Philosophy at University of Illinois, introduced the formalism of graphical causal models and how to interpret them, discussed some results in causal search, as well as the problem of mixed populations.
Lucia Petito, Assistant Professor of Preventive Medicine (Biostatistics) at Northwestern University, introduced participants to the estimation of intention-to-treat and per-protocol effects in randomized trials with survival outcomes, and how to extend these concepts to research done in “found” data using the target trial concept.
Jacob Bowers, Associate Professor of Political Science and Statistics at University of Illinois, showed how a testing-based approach to causal inference can be used to complement estimation-based approaches in complex but common situations. This can include when an experiment occurs in thousands of sites, and policymakers’ interest may lie in detecting effects in specific sites rather than in estimating an average effect within each site.
Felix Thoemmes, Associate Professor of Human Development at Cornell University, discussed the history of causal thinking in the field of psychology, juxtaposed with developments outside the field, and how current examples from psychological research could be improved with considerations of causal effects.
Rodrigo Pinto, Assistant Professor of Economics at the University of California Los Angeles, addressed economic incentives, human behavior, and the design of social experiments and showed how instrumental variables can be applied to examine some important social experiments that fight poverty.
Powerpoint slides of each presentation are also available to view and to download.