Why You Are Using PROC GLM Too Much (and What You Should Be Using Instead) Part 2

Presented: Friday June 18, 2021, 10:00am-2:00pm Pacific Daylight Time

Presented by:
Deanna Schreiber-Gregory is a Lead Research Statistician and Data Manager on contract through the Henry M Jackson Foundation for the Advancement of Military Medicine to the Department of Defense in Bethesda, MD. She is also an Independent Consultant for Statistics, Research Methods, and Data Management in the private sector through Juxdapoze, LLC. Deanna has an MS in Health and Life Science Analytics, a BS in Statistics, and a BS in Psychology. Deanna has presented as a contributed and invited speaker at over 50 local, regional, national, and global SAS user group conferences since 2011.

Peter Flom is a retired independent statistical consultant who worked with graduate students and researchers in the social, medical and behavioral sciences. He has been using SAS for over 20 years and has given talks at SGF and many local and regional groups.

Description:The general linear model (linear regression and ANOVA) is one of the most commonly used statistical methods. However, the GLM makes assumptions and sometimes these assumptions are violated. There are many techniques that can be used to deal with various violations, and there are SAS PROCs to implement these. These include: Quantile regression, Robust regression, Cubic splines and other forms of splines, Multivariate adaptive regression splines (MARS), Regression trees, Multilevel models, Ridge Regression, LASSO, and Elastic Nets, among other methods. Covered PROCs include QUANTREG, ROBUSTREG, ADAPTIVEREG and MIXED.

Part 2: Ridge regression (REG), Lasso and elastic nets (GLMSELECT), and multilevel models (MIXED and GLIMMIX).