Abstract #300730

This is the preliminary program for the 2003 Joint Statistical Meetings in San Francisco, California. Currently included in this program is the "technical" program, schedule of invited, topic contributed, regular contributed and poster sessions; Continuing Education courses (August 2-5, 2003); and Committee and Business Meetings. This on-line program will be updated frequently to reflect the most current revisions.

To View the Program:
You may choose to view all activities of the program or just parts of it at any one time. All activities are arranged by date and time.

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

Back to main JSM 2003 Program page
JSM 2003 Abstract #300730
Activity Number: 216
Type: Contributed
Date/Time: Tuesday, August 5, 2003 : 8:30 AM to 10:20 AM
Sponsor: Biometrics Section
Abstract - #300730
Title: Diagnosing Non-Normality of Random Effects in General Linear Mixed Models: Can We Distinguish between an Incorrect Mean Model and an Incorrect Random Effects Model?
Author(s): Lynn E. Eberly*+ and Lisa M. Thackeray
Companies: University of Minnesota and University of Minnesota
Address: Box 303 UMHC, Minneapolis, MN, 55455-0341,
Keywords: longitudinal data ; REML ; diagnostic ; normal probability plot ; quantile plot
Abstract:

Random effects models are useful tools for longitudinally collected continuous data, but most computer implementations of these models require the assumption of normality for the random effects. Ryan and Lange (1989) describe a normal quantile plot to be used as a diagnostic for the normality of the random effects. An estimator is constructed from a standardized linear combination of the estimated random effects. Quantile plots and standard error bars can be computed for several linear combinations and compared. We consider a model with random intercepts and slopes and show with simulated data that the plots do show indications of model inadequacy when a model assumption is violated. However, they often cannot distinguish between an incorrectly specified mean model, such as a quadratic effect modeled as linear, and an incorrectly specified random effects distribution, such as skewed or bimodal. The plots are also more sensitive to non-normality of random intercepts, and less sensitive to non-normality of random slopes.

  • The address information is for the authors that have a + after their name.
  • Authors who are presenting talks have a * after their name.

Back to the full JSM 2003 program

JSM 2003 For information, contact meetings@amstat.org or phone (703) 684-1221. If you have questions about the Continuing Education program, please contact the Education Department.
Revised March 2003