**
George J. Knafl, Ph.D.
**

SAS® Software for Conducting a Variety of Analyses

These macros can be loaded into SAS with %include statements. Use is limited as indicated in the documentation in the code. For questions, contact gknafl@unc.edu. All rights reserved. No warranties granted, explicit or implicit. These are the most recent versions. Versions available with the examples have not been updated so that they may be used to reproduce results reported for those examples. Requires Base SAS, SAS/STAT, and PROC IML. Development of these tools has been supported in part by Grant # R01 AI57043 from the National Institute of Allergy and Infectious Diseases and Grant # R03 MH086132 from the National Institute of Mental Health.

Internet Explorer does not always download SAS code and data files correctly. To get around this, right click on the file and use the Save target as option.

- SAS macro compmed.20160924.sas for (moderated) monotonic mediation modeling. Documentation for the compmed macro.
- SAS macro genclass.20170901.sas for adaptive classification modeling. Documentation for the genclass macro.
- SAS macro genreg.20170901.sas for adaptive regression modeling. See also adaptive regression examples. Documentation for the genreg macro.
- SAS macro grpevnts.20170807.sas for grouping electronic adherence data. See also electronic adherence data analysis examples. Documentation for the grpevnts macro.
- SAS macro impute.20141005.sas for imputing item values. See also composite reliability examples.
- SAS macro LCVCFA.20101001.sas for confirmatory factor analysis. Note that this macro has been replaced by the more general LCVSEM macro. See also factor analysis examples.
- SAS macro LCVcluster.20170827.sas for adaptive cluster analysis. See also electronic adherence data analysis examples. Documentation for the LCVcluster macro.
- SAS macro LCVEFA.20101002.sas for exploratory factor analysis. See also factor analysis examples.
- SAS macro LCVGAM.20150910.sas for generalized additive modeling. Requires SAS Version 9.3 or later. See also adaptive fractional polynomial modeling. Documentation for the LCVGAM macro.
- SAS macro LCVmixed.20140810.sas for linear mixed modeling. See also analyzing longitudinal data.
- SAS macro LCVSEM.20141115.sas for structural equation modeling. Documentation for the LCVSEM macro.
- SAS macro MARSmodl.20150910.sas for generating multiple adaptive regression spline (MARS) models to be able to compute their LCV scores with the genreg macro Documentation for the MARSmodl macro.
- SAS macro medCIs.20170807.sas for bootstrapped confidence intervals for indirect effects due to mediation. Use medCIs only for linear indirect effects. Although it supports nonlinear indirect effects, its use in this case is not recommended. Use the compmed macro instead. Documentation for the medCIs macro.
- SAS macro modmedCIs.20160501.sas for bootstrapped confidence intervals for nonlinear indirect effects due to mediation. See also mediation analysis examples. Documentation for the modmedCIs macro.
- SAS macros multGAM.20150320.sas for adaptive generalized additive modeling. Requires SAS Version 9.3 or later. See also adaptive fractional polynomial modeling. Documentation for the multGAM macros.
- SAS macro multsubj.20170807.sas for analysis of electronic adherence data for multiple subjects possibly within multiple phases. Replaces older versions of both the multsubj and multphse macros. See also electronic adherence data analysis examples. Documentation for the multsubj macro.
- SAS macro RAmacros.20170806.sas for assessing Royston and Altman recommended fractional polynomial models. See also adaptive fractional polynomial modeling. Documentation for the RAmacros.
- SAS macros reliability.20140426.sas for computing composite reliability and construct validity coefficients for family data. See also composite reliability examples.
- SAS macro ypower.20170806.sas for power transforming positive valued continuous outcomes. Documentation for the ypower macro.