|
Abstract:
|
This workshop introduces two SAS/STAT® procedures, CAUSALRT and PSMATCH, for the analysis of causal treatment effects from observational data. In this setting, confounding variables, which are associated with both treatment assignment and outcomes, can bias estimates of causal treatment effects if not dealt with carefully. The CAUSALTRT and PSMATCH procedures, which are new in SAS/STAT 14.2, implement methods that help you deal with confounding. The CAUSALTRT procedure estimates treatment effects by propensity score weighting, regression adjustment, or a combination of the two (the doubly-robust method). The PSMATCH procedure employs the propensity score model to create matched samples that behave like data from a randomized experiment. Alternatively, the PSMATCH procedure can create data sets that you can use to estimate causal effects by propensity score weighting or stratification. This workshop demonstrates the uses of these two procedures through examples. It also gives a brief, high-level account of causal inference issues and the principles that underlie the two procedures. Basic familiarity with generalized linear models is assumed.
|
