Gregg, M.
, Creuziger, A.
, Datta, S.
and Lorenz, D.
(2024),
Two tests of variance homogeneity for clustered data where group size is informative, Journal of Statistical Computation and Simulation, [online], https://doi.org/10.1080/00949655.2024.2430692, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=957607
(Accessed January 13, 2025)
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
Secure .gov websites use HTTPS
A lock (
) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.
Two tests of variance homogeneity for clustered data where group size is informative
Published
Author(s)
, , Somnath Datta, Douglas Lorenz
Abstract
To evaluate variance homogeneity among groups for clustered data, Iachine et al. (Robust tests for the equality of variances for clustered data. J Stat Comput Simul 2010;80(4):365–377) introduced an extension of the well-known Levene test. However, this method does not account for informative cluster size (ICS) or informative within-cluster group size (IWCGS), which can occur in clustered data when cluster and group sizes are random variables. This article introduces two tests of variance homogeneity that are appropriate for data with ICS and IWCGS, one extending the Levene-style transformation method and one based on a direct comparison of estimates of variance. We demonstrate the properties of our tests in a detailed simulation study and show that they are resistant to the potentially biasing effects of ICS and IWCGS. We illustrate the use of these tests by applying them to a data set of x-ray diffraction measurements collected from a specimen of duplex steel.Citation
Journal of Statistical Computation and Simulation
Pub Type
Journals
Download Paper
Keywords
Homogeneity of variance, robust tests, clustered data, informative cluster size, informative within-cluster group size
Citation
Issues
If you have any questions about this publication or are having problems accessing it, please contact reflib@nist.gov.
Created December 3, 2024, Updated December 5, 2024