Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 521 - Advances in Methods and Novel Applications for Data with Hierarchies and Under/Over-Dispersion
Type: Invited
Date/Time: Thursday, August 6, 2020 : 1:00 PM to 2:50 PM
Sponsor: Section on Statistics in Epidemiology
Abstract #309196
Title: Meta-Analysis of Summary Statistics from K X M Tables Accounting for Over-Dispersion
Author(s): Majnu John* and Heejung Bang and Victoria Landsman
Companies: Northwell Health/Hofstra and Division of Biostatistics, University of California and University of Toronto
Keywords: meta analysis; overdispersion; kappa statistic; diagnostic odds ratio; Dirichlet-Multinomial

Summary statistics such as diagnostic odds ratio or kappa statistic are derived from K x M tables which could be modeled using multinomial distribution. We present Dirichlet-Multinomial models to account for overdispersion in such scenarios. In a meta-analysis context we propose two methods under this modeling scenario, one based on Generalized Estimating Equations and the other a Mantel-Haenszel type method, to obtain a pooled estimate. We study the advantages and limitations of the two methods using simulations and illustrate the methods using real data analysis.

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

Back to the full JSM 2020 program