Hello, everyone
Today I read a paper which used the PS Fine Stratification as their statistical analysis method. It was inspiring as the IPTW method can inflate the pseudo-cohort by outliers (by extreme low/high propensity scores). I know the robust variance estimator can handle the correlation between clone observations but the inflation itself can affect the accuracy of the confidence intervals.
By the method of PS Fine Stratification, the IPTW-based inflation is not a problem as it will not use a pseudo-cohort. Therefore, I wonder if we can do PS Fine Stratification in SAS by existing syntaxes. Thank you.
