Date on Master's Thesis/Doctoral Dissertation


Document Type

Master's Thesis

Degree Name



Bioinformatics and Biostatistics

Degree Program

Biostatistics, MS

Committee Chair

Kong, Maiying

Committee Co-Chair (if applicable)

Benitez, Joseph

Committee Member

Benitez, Joseph

Committee Member

Gaskins, Jeremy

Author's Keywords

propensity scores; causal inference; observational data; cardiotoxicity; breast cancer; average treatment effect


Observational data presents unique challenges for analysis that are not encountered with experimental data resulting from carefully designed randomized controlled trials. Selection bias and unbalanced treatment assignments can obscure estimations of treatment effects, making the process of causal inference from observational data highly problematic. In 1983, Paul Rosenbaum and Donald Rubin formalized an approach for analyzing observational data that adjusts treatment effect estimates for the set of non-treatment variables that are measured at baseline. The propensity score is the conditional probability of assignment to a treatment group given the covariates. Using this score, one may balance the covariates across treatment groups and obtain unbiased estimates of treatment effects. This paper has three objectives: to explain propensity scores, their assumptions, and their applications; to illustrate their use and several considerations underlying various propensity score methods, by using simulation studies; and to use propensity score methods to estimate average treatment effect between two types of breast cancer chemotherapy treatment regimens, with respect to subsequent development of cardiotoxicity.