Department of Biostatistics

People

Prospective Students

Academics

Research

News & Events

Calendar

Consulting

Employment Opportunities

Resource Quicklinks

Computing Environment

Contact

THESIS DEFENSE ABSTRACT

Counterfactual Inference from Observational Data: Methods and Applications

Weiwei Wang, PhD Candidate, Johns Hopkins Department of Biostatistics

This dissertation discusses counterfactual, statistical inference in three observational data settings.  In the first setting, we seek to estimate, in two separate studies of patients with acute lung injury, the survival distribution in the counterfactual world in which patients had adhered with a time-varying dynamic treatment regime and had their observational data fully recorded.  In the second setting, we seek to estimate, from an outcome-dependent sample of moderately-severe injured trauma patients, the probability of in-hospital mortality in two counterfactual worlds: one in which they were all treated at trauma centers and one in which they were all treated at non-trauma centers.  In the third setting, we seek to estimate, from an observational cohort of Baltimore city residents, the age-specific incidence proportion of depression in the counterfactual world in which all individuals without a history of depression at the baseline interview provided information about their depression status through the minimum of their scheduled follow-up interview or death.