Article Version of Record

Recovering crossed random effects in mixed-effects models using model averaging

Author(s) / Creator(s)

Martínez-Huertas, José Ángel
Olmos, Ricardo

Abstract / Description

Random effects contain crucial information to understand the variability of the processes under study in mixed-effects models with crossed random effects (MEMs-CR). Given that model selection makes all-or-nothing decisions regarding to the inclusion of model parameters, we evaluated if model averaging could deal with model uncertainty to recover random effects of MEMs-CR. Specifically, we analyzed the bias and the root mean squared error (RMSE) of the estimations of the variances of random effects using model averaging with Akaike weights and Bayesian model averaging with BIC posterior probabilities, comparing them with two alternative analytical strategies as benchmarks: AIC and BIC model selection, and fitting a full random structure. A simulation study was conducted manipulating sample sizes for subjects and items, and the variance of random effects. Results showed that model averaging, especially Akaike weights, can adequately recover random variances, given a minimum sample size in the modeled clusters. Thus, we endorse using model averaging to deal with model uncertainty in MEMs-CR. An empirical illustration is provided to ease the usability of model averaging.

Keyword(s)

mixed-effects models crossed random effects random effects model averaging Akaike weights Bayesian model averaging AIC BIC

Persistent Identifier

Date of first publication

2022-12-23

Journal title

Methodology

Volume

18

Issue

4

Page numbers

298–323

Publisher

PsychOpen GOLD

Publication status

publishedVersion

Review status

peerReviewed

Is version of

Citation

Martínez-Huertas, J., & Olmos, R. (2022). Recovering crossed random effects in mixed-effects models using model averaging. Methodology, 18(4), 298-323. https://doi.org/10.5964/meth.9597
  • Author(s) / Creator(s)
    Martínez-Huertas, José Ángel
  • Author(s) / Creator(s)
    Olmos, Ricardo
  • PsychArchives acquisition timestamp
    2023-01-23T14:06:47Z
  • Made available on
    2023-01-23T14:06:47Z
  • Date of first publication
    2022-12-23
  • Abstract / Description
    Random effects contain crucial information to understand the variability of the processes under study in mixed-effects models with crossed random effects (MEMs-CR). Given that model selection makes all-or-nothing decisions regarding to the inclusion of model parameters, we evaluated if model averaging could deal with model uncertainty to recover random effects of MEMs-CR. Specifically, we analyzed the bias and the root mean squared error (RMSE) of the estimations of the variances of random effects using model averaging with Akaike weights and Bayesian model averaging with BIC posterior probabilities, comparing them with two alternative analytical strategies as benchmarks: AIC and BIC model selection, and fitting a full random structure. A simulation study was conducted manipulating sample sizes for subjects and items, and the variance of random effects. Results showed that model averaging, especially Akaike weights, can adequately recover random variances, given a minimum sample size in the modeled clusters. Thus, we endorse using model averaging to deal with model uncertainty in MEMs-CR. An empirical illustration is provided to ease the usability of model averaging.
    en_US
  • Publication status
    publishedVersion
  • Review status
    peerReviewed
  • Citation
    Martínez-Huertas, J., & Olmos, R. (2022). Recovering crossed random effects in mixed-effects models using model averaging. Methodology, 18(4), 298-323. https://doi.org/10.5964/meth.9597
    en_US
  • ISSN
    1614-2241
  • Persistent Identifier
    https://hdl.handle.net/20.500.12034/7990
  • Persistent Identifier
    https://doi.org/10.23668/psycharchives.12449
  • Language of content
    eng
  • Publisher
    PsychOpen GOLD
  • Is version of
    https://doi.org/10.5964/meth.9597
  • Keyword(s)
    mixed-effects models
    en_US
  • Keyword(s)
    crossed random effects
    en_US
  • Keyword(s)
    random effects
    en_US
  • Keyword(s)
    model averaging
    en_US
  • Keyword(s)
    Akaike weights
    en_US
  • Keyword(s)
    Bayesian model averaging
    en_US
  • Keyword(s)
    AIC
    en_US
  • Keyword(s)
    BIC
    en_US
  • Dewey Decimal Classification number(s)
    150
  • Title
    Recovering crossed random effects in mixed-effects models using model averaging
    en_US
  • DRO type
    article
  • Issue
    4
  • Journal title
    Methodology
  • Page numbers
    298–323
  • Volume
    18
  • Visible tag(s)
    Version of Record
    en_US