Level: Beginner
In observational studies and epidemiological studies that make comparisons between groups to estimate treatment effects, imbalances in subject background factors often occur. Imbalances between groups in potentially confounding background factors can impair comparability, so careful analysis is required. Generally, imbalances in confounding factors can be adjusted to some extent by stratified analysis, but stratified analysis may not be applicable when there are many confounding factors compared to the sample size.
One of the adjustment methods used in such cases is analysis using propensity scores. Propensity scores are primarily used for stratified analysis, matching, and inverse probability weighting (IPW), and matching can be implemented using a JMP add-in. However, stratified analysis using propensity scores and IPW are not implemented in JMP, and several manual steps are required. Because multiple propensity score models are considered in analyses using propensity scores, automating the steps is important for improving research efficiency.
This presentation will focus on IPW using propensity scores, and will demonstrate how to run it in JMP and how to automate the procedure.
