Weighted GEE Analysis Using SAS/STAT Software (ADDED FEE)

• Activity Number: 214610 - Weighted GEE Analysis Using SAS/STAT Software (ADDED FEE)
• Type: Professional Development
• Date/Time: Wednesday, August 2, 2017 : 3:00 PM to 4:45 PM
• Sponsor: ASA
• Author(s): Michael Lamm*
• Companies: SAS Institute Inc.

Abstract

The generalized estimating equation (GEE) approach is commonly used for the analysis of longitudinal data. The standard GEE approach is based on complete cases and can lead to biased results when observations are missing due to dropouts or skipped visits. When your data include dropouts, weighted GEE methods can be used to eliminate bias. This workshop introduces you to the GEE procedure (new in the SAS/STAT 13.2 release) which supports both the standard and weighted GEE methods for analyzing longitudinal data. In this workshop you will learn about the different mechanisms used to describe why a response is missing, and how the missing data mechanism impacts inference using the standard and weighted GEE approaches. This workshop illustrates the use of PROC GEE with examples, and compares the methods available in PROC GENMOD and PROC GEE. A basic familiarity with generalized linear models is assumed.