Article Version of Record

Parametric and semi-parametric bootstrap-based confidence intervals for robust linear mixed models

Author(s) / Creator(s)

Mason, Fabio
Cantoni, Eva
Ghisletta, Paolo

Abstract / Description

The linear mixed model (LMM) is a popular statistical model for the analysis of longitudinal data. However, the robust estimation of and inferential conclusions for the LMM in the presence of outliers (i.e., observations with very low probability of occurrence under Normality) is not part of mainstream longitudinal data analysis. In this work, we compared the coverage rates of confidence intervals (CIs) based on two bootstrap methods, applied to three robust estimation methods. We carried out a simulation experiment to compare CIs under three different conditions: data 1) without contamination, 2) contaminated by within-, or 3) between-participant outliers. Results showed that the semi-parametric bootstrap associated to the composite tau-estimator leads to valid inferential decisions with both uncontaminated and contaminated data. This being the most comprehensive study of CIs applied to robust estimators of the LMM, we provide fully commented R code for all methods applied to a popular example.

Keyword(s)

robustness linear mixed models bootstrap confidence intervals longitudinal data

Persistent Identifier

Date of first publication

2021-12-17

Journal title

Methodology

Volume

17

Issue

4

Page numbers

271–295

Publisher

PsychOpen GOLD

Publication status

publishedVersion

Review status

peerReviewed

Is version of

Citation

Mason, F., Cantoni, E., & Ghisletta, P. (2021). Parametric and semi-parametric bootstrap-based confidence intervals for robust linear mixed models. Methodology, 17(4), 271-295. https://doi.org/10.5964/meth.6607
  • Author(s) / Creator(s)
    Mason, Fabio
  • Author(s) / Creator(s)
    Cantoni, Eva
  • Author(s) / Creator(s)
    Ghisletta, Paolo
  • PsychArchives acquisition timestamp
    2022-04-14T11:24:57Z
  • Made available on
    2022-04-14T11:24:57Z
  • Date of first publication
    2021-12-17
  • Abstract / Description
    The linear mixed model (LMM) is a popular statistical model for the analysis of longitudinal data. However, the robust estimation of and inferential conclusions for the LMM in the presence of outliers (i.e., observations with very low probability of occurrence under Normality) is not part of mainstream longitudinal data analysis. In this work, we compared the coverage rates of confidence intervals (CIs) based on two bootstrap methods, applied to three robust estimation methods. We carried out a simulation experiment to compare CIs under three different conditions: data 1) without contamination, 2) contaminated by within-, or 3) between-participant outliers. Results showed that the semi-parametric bootstrap associated to the composite tau-estimator leads to valid inferential decisions with both uncontaminated and contaminated data. This being the most comprehensive study of CIs applied to robust estimators of the LMM, we provide fully commented R code for all methods applied to a popular example.
    en_US
  • Publication status
    publishedVersion
  • Review status
    peerReviewed
  • Citation
    Mason, F., Cantoni, E., & Ghisletta, P. (2021). Parametric and semi-parametric bootstrap-based confidence intervals for robust linear mixed models. Methodology, 17(4), 271-295. https://doi.org/10.5964/meth.6607
    en_US
  • ISSN
    1614-2241
  • Persistent Identifier
    https://hdl.handle.net/20.500.12034/5710
  • Persistent Identifier
    https://doi.org/10.23668/psycharchives.6314
  • Language of content
    eng
  • Publisher
    PsychOpen GOLD
  • Is version of
    https://doi.org/10.5964/meth.6607
  • Is related to
    https://doi.org/10.23668/psycharchives.5302
  • Is related to
    https://github.com/masonFG/CIrobustLMM
  • Is related to
    https://doi.org/10.23668/psycharchives.6322
  • Keyword(s)
    robustness
    en_US
  • Keyword(s)
    linear mixed models
    en_US
  • Keyword(s)
    bootstrap
    en_US
  • Keyword(s)
    confidence intervals
    en_US
  • Keyword(s)
    longitudinal data
    en_US
  • Dewey Decimal Classification number(s)
    150
  • Title
    Parametric and semi-parametric bootstrap-based confidence intervals for robust linear mixed models
    en_US
  • DRO type
    article
  • Issue
    4
  • Journal title
    Methodology
  • Page numbers
    271–295
  • Volume
    17
  • Visible tag(s)
    Version of Record
    en_US