Biostatistics Weekly Seminar

Weighting methods are widely used to adjust for covariates in observational studies, sample surveys, and regression settings. In this talk, we will study a class of recently proposed weighting methods that, instead of modeling the probabilities of missingness or treatment, they focus on balancing the observed covariates with weights of minimum dispersion. We will call these weights 'minimal weights' and study them under a common optimization framework, both from theoretical and practical standpoints. From a theoretical standpoint, we will characterize the asymptotic properties of minimal weights and show that under standard smoothness conditions on the propensity score function minimal weights are consistent estimates of the true inverse probability weights. Also, we will show that the resulting weighting estimator is consistent, asymptotically normal, and semiparametrically efficient. From a practical standpoint, we will provide a tuning algorithm for choosing the degree of approximate balance in minimal weights, which can be of independent interest. Also, we will discuss empirical studies that suggest approximate balance is preferable to exact balance, especially when there is limited overlap in covariate distributions. In these studies, we will show that the root mean squared error of the weighting estimator can be reduced by as much as a half with approximate balance. Finally, we will discuss extensions to settings with multivalued and time-varying treatments.

MRB III 1220
29 August 2018
1:30pm