Online Program Home
My Program

Abstract Details

Activity Number: 562
Type: Contributed
Date/Time: Wednesday, August 3, 2016 : 11:35 AM to 12:20 PM
Sponsor: Biometrics Section
Abstract #321797
Title: The Impact of Model Misspecification on Repeated Measures Analysis
Author(s): Yingmei Xi* and John Jiang
Companies: Vertex Pharmaceuticals and Vertex Pharmaceuticals
Keywords: MMRM ; Within-subject correlation matrix ; Proc Mixed procedure ; Simulation

The mixed-effects model for repeated measures (MMRM) is widely used in clinic trials. In SAS, the PROC Mixed procedure provides many different choices of within-subject correlation matrix. This simulation study is conducted in order to better understand how MMRM behaves when using different correlation matrixes. The data are simulated under a few correlation structures. We analyze the data by PROC MIXED with some commonly used factors as well as different choices of correlation matrices. Then we compare the type 1 error rates, biases, standard errors and powers for all the analyses. In addition, we try to understand if the sample size has any impact on the analysis results under different scenarios. At the same time, we evaluate if the convergence is a problem for the study models. At the end, we conclude that the within-subject correlation structure must be appropriately selected in order to draw a correct conclusion from the repeated measures data.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association