Multilevel models are commonly used in research designs where experimental or sampling units are organized at more than one level. Data with a hierarchical or nested structure are commonly seen in clinical and educational research designs.
SAS PROC MIXED is a procedure that fits a variety of mixed linear models to data and enables you to use these fitted models to make statistical inferences about the data. If you are new to multilevel modeling, or to using SAS PROC MIXED, the following paper provides a great introduction to the topic and includes numerous examples that provide the research questions being examined, the code used to fit the models, as well as how to interpret the output.
Dr. Bethany Bell and two of her graduate students from the University of South Carolina presented their paper as part of a livestream session at SAS Global Forum 2013. The paper and video are two great resources to help you get started using SAS PROC MIXED.
“A Multilevel Model Primer Using SAS PROC MIXED” by Bethany A. Bell, Mihaela Ene, Whitney Smiley, and Jason A. Schoenberger.
Another great resource for multilevel modeling is Multilevel Modeling of Hierarchical and Longitudinal Data Using SAS and for PROC MIXED in general Mixed Models Analyses Using SAS.
For more resources related to Statistics, visit the International Year of Statistics website where you’ll find tips and techniques, as well as featured articles related to the field of Statistics.
University Program Spotlight
Dr. Bethany Bell is an Assistant Professor in the Department of Educational Studies at the University of South Carolina. Dr. Bell is a 2013 SAS Global Forum Faculty Scholarship recipient. In addition to being recognized at SAS Global Forum, that same week Dr. Bell received the 2013 AERA Early Career Award for Quantitative Methods and Statistical Theory. Two of Dr. Bell’s graduate students also received awards at SAS Global Forum. Whitney Smiley is a 2013 SAS Student Ambassador and Mihaela Ene is a 2013 SAS Student Ambassador Honorable Mention award winner.