ABSTRACT
This chapter introduces how to implement the proposed methods in Chapter 4 using the common statistical softwares: R and SAS. An R package mma was complied. Using the package, the researcher can perform the third-variable analysis step by step: first, identify potential third-variables, and then perform the analysis to get estimates of third-variable effects based on the whole dataset, and finally using bootstrap method to estimate the variances and confidence intervals for the estimates. All steps can also be implemented in one combined function. We then introduce how to call the R package and obtain the third-variable effect estimates within the SAS environment. SAS macros are provided for the implementation. Finally, we perform the third-variable analysis to explore the racial disparity in breast cancer survival using both linear and nonlinear third-variable analysis. In addition, we perform a series of simulations to check the estimation precisions and power of the proposed method.
