Introduction to propensity score methods
Date
2021
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Randomised controlled trials (RCTs), while the golden standard of estimating causal effects in clinical studies, are not always possible to conduct due to ethical reasons or other restrictions. Observational studies are an alternative in such cases. However, in such studies, treatment assignment may be subject to systematic biases.
Propensity score (PS) methods are a popular tool to adjust for confounding factors in observational studies. By attempting to mimic RCTs, these methods are quite intuitive. This thesis provides a theoretical overview of the most popular PS methods, and conducts a simulation study to compare PS matching, PS weighting, and conventional covariate adjustment. CERCS research specialisation: P160 Statistics, operation research,
Description
Keywords
kalduvuse määr, vaatlusuuringud, ühildamine, kaalumine, tunnustele kohandamine, covariate adjustment, weighting, matching, observational studies, propensity score