Methods - Pseudo-simulation Sample Clauses

Methods - Pseudo-simulation. As mentioned in section 3.1, this first round of simulations is based on a re-sampling of existing data to create datasets with known covariate distributions of interest but unknown treatment effect sizes. Based on the pool of variables described above, propensity scores were calculated using logistic regression with race as the outcome. Possible covariates con- sidered for this model included basic demographic and medical history information (age, residence, mode of arrival to emergency department (ED), diabetes, etc.), several of which were significantly associated with race (i.e., imbalanced between the two groups), see table