Here is the outline of the 3-hour workshop. Here are the workshop slides.

The methodology discussed at the workshop is written about in these two articles:

- Nahum-Shani I, Qian M, Almirall D, Pelham WE, Gnagy B, Waxmonsky J, Yu J, Murphy SA. (submitted)
*Experimental Design and Primary Data Analysis Methods for Comparing Adaptive Interventions.*Technical Report, Methodology Center, Pennsylvania State University. - Nahum-Shani I, Qian M, Almirall D, Pelham WE, Gnagy B, Waxmonsky J, Yu J, Murphy SA. (submitted)
*Q-Learning: A Data Analysis Method for Constructing Adaptive Interventions.*Technical Report, Methodology Center, Pennsylvania State University. - Chakraborty B, Murphy SA, and Strecher V (2009)
*Inference for non-regular parameters in optimal dynamic treatment regimes.*Statistical Methods in Medical Research, 19(3), 317-343.
Statistical Methods in Medical Research, 19(3), 317-343.

Here is the simulated/fake data set, in .SAS7BDAT format, that is used with the following SAS code. Here is the SAS code accompanying the slides in Module 3 (starting on page 60) of the workshop. Here is the SAS code accompanying the slides in Module 4 (starting on page 92) of the workshop.

Here is the (same as above) simulated data set (but in .TXT format) used with the following R code. Here is the R code that produces the results and graphs related to Q-Learning Regression shown in the slides in Module 4 (starting on page 105) of the workshop. To use this R code you must first install R, then you must install the Q-Learning package given here. To install from within R, click Packages, then click Install Package(s) from local zip files. Once installed, type the command library(qlearning) in R to load the package (now you are ready to use the R code above). The Q-Learning methodology that is coded into this software is based on the work in Chakraborty et al. (2009, full citation above).

IMPORTANT: Stay tuned to the Methodology Center website (click on Free Software/Appelts), as SAS code (with updated inferential methods) will soon be available to perform Q-Learning regression for developing optimal adaptive treatment strategies. This webpage may not be revised even when the new software becomes available.

You should first run the SAS and R code above and see if you arrive at the same (simulated data) results presented in the workshop slides. Then you can tweak the SAS and R code for your own purposes and using your own data.

R is a language and environment for statistical computing and graphics. You will need to install R and then install the Q-learning package in order to use the R code above. R can be downloaded for free by visiting the R-Project Home Page.

If you use any of this work in your own research, we would greatly appreciate it if you cite the articles listed at the top of this web page. Thank you very much!

Funding for this research was provided by the following grants: R01-MH-080015 (Murphy), and P50-DA-010075 (Murphy).

**First Published:** 4/29/2011; **Last Revised**: 4/29/2011