The influence of misspecification of the heteroscedasticity on multilevel regression parameter and standard error estimates
The influence of misspecification of the heteroscedasticity on multilevel regression parameter and standard error estimates
Like in ordinary regression models, in multilevel analysis, homoscedasticity of the residual variances is an assumption that is mostly unchecked. However, in experimental research, the residual variance component at level two may differ in the experimental and the control condition, leading to heteroscedastic second level variances. Using a simulation study, the consequences of ignoring second level heteroscedasticity on the estimation of the fixed and random parameters and their standard errors was investigated. It was found that the standard error of the second level variance is underestimated, but that the estimated fixed parameters of the independent variables, the first level variance and their standard errors are mostly unbiased.
heteroscedasticity, homoscedasticity, heterogeneous variances, multilevel analysis, cluster randomized trials, simulation
67-72
van der Heijden, Peter G.M.
85157917-3b33-4683-81be-713f987fd612
Moerbeek, Mirjam
cb334a2e-5e73-450d-855d-5655f285e221
Maas, Cora J.M.
d70333f4-890c-4b59-b13f-7fbec656ad06
Korendijk, Elly J.H.
c335f168-448c-4fc8-9ff3-88d64f978d94
2008
van der Heijden, Peter G.M.
85157917-3b33-4683-81be-713f987fd612
Moerbeek, Mirjam
cb334a2e-5e73-450d-855d-5655f285e221
Maas, Cora J.M.
d70333f4-890c-4b59-b13f-7fbec656ad06
Korendijk, Elly J.H.
c335f168-448c-4fc8-9ff3-88d64f978d94
van der Heijden, Peter G.M., Moerbeek, Mirjam, Maas, Cora J.M. and Korendijk, Elly J.H.
(2008)
The influence of misspecification of the heteroscedasticity on multilevel regression parameter and standard error estimates.
Methodology: European Journal of Research Methods for the Behavioral and Social Sciences, 4 (2), .
(doi:10.1027/1614-2241.4.2.67).
Abstract
Like in ordinary regression models, in multilevel analysis, homoscedasticity of the residual variances is an assumption that is mostly unchecked. However, in experimental research, the residual variance component at level two may differ in the experimental and the control condition, leading to heteroscedastic second level variances. Using a simulation study, the consequences of ignoring second level heteroscedasticity on the estimation of the fixed and random parameters and their standard errors was investigated. It was found that the standard error of the second level variance is underestimated, but that the estimated fixed parameters of the independent variables, the first level variance and their standard errors are mostly unbiased.
This record has no associated files available for download.
More information
Published date: 2008
Keywords:
heteroscedasticity, homoscedasticity, heterogeneous variances, multilevel analysis, cluster randomized trials, simulation
Organisations:
Statistical Sciences Research Institute
Identifiers
Local EPrints ID: 344676
URI: http://eprints.soton.ac.uk/id/eprint/344676
ISSN: 1614-1881
PURE UUID: 563198fb-4a6e-448f-b94f-310ca14fcfab
Catalogue record
Date deposited: 26 Oct 2012 13:53
Last modified: 15 Mar 2024 03:46
Export record
Altmetrics
Contributors
Author:
Mirjam Moerbeek
Author:
Cora J.M. Maas
Author:
Elly J.H. Korendijk
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
