Online Program Home
My Program

Abstract Details

Activity Number: 495
Type: Contributed
Date/Time: Wednesday, August 3, 2016 : 8:30 AM to 10:20 AM
Sponsor: Biometrics Section
Abstract #319542
Title: Confidence Intervals for Variance Components in Mixed-Effects Models: A Likelihood-Based Approach
Author(s): Gaurav Sharma* and Lihan Yan
Companies: The EMMES Corporation and FDA
Keywords: Bioassay ; Variance Components ; Confidence Interval ; Likelihood ; Coverage Probability ; Vaccine

Biological Assay (bioassay) plays a major role in assessing the biological activities in response to biological drug products such as vaccines. Characterization of the precision of a bioassay is a fundamental element during the validation of a bioassay. The precision is often estimated through the variance components which could arise due to various factors such as runs, days, analysts, and replicates. Mixed-effects models are used to obtain the confidence intervals for these variance components. Extensive simulation results show that in some cases the confidence intervals commonly obtained using the Wald's approach (implemented in SAS) exhibit poor coverage probability. In this presentation, we propose a likelihood-based approach to obtain the confidence intervals for variance components in the context of the general mixed effects models. Numerical simulation results show that the confidence intervals derived based on such an approach perform better than the competing confidence interval procedures from the literature.

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

Back to the full JSM 2016 program

Copyright © American Statistical Association