Cluster-robust variance estimation for dyadic data

Peter M. Aronow, Cyrus Samii, Valentina A. Assenova

    Research output: Contribution to journalArticle

    Abstract

    Dyadic data are common in the social sciences, although inference for such settings involves accounting for a complex clustering structure. Many analyses in the social sciences fail to account for the fact that multiple dyads share a member, and that errors are thus likely correlated across these dyads. We propose a non-parametric, sandwich-type robust variance estimator for linear regression to account for such clustering in dyadic data. We enumerate conditions for estimator consistency. We also extend our results to repeated and weighted observations, including directed dyads and longitudinal data, and provide an implementation for generalized linear models such as logistic regression. We examine empirical performance with simulations and an application to interstate disputes.

    Original languageEnglish (US)
    Article numbermpv018
    Pages (from-to)564-577
    Number of pages14
    JournalPolitical Analysis
    Volume23
    Issue number4
    DOIs
    StatePublished - Oct 1 2015

    Fingerprint

    dyad
    social science
    regression
    linear model
    logistics
    simulation
    performance

    ASJC Scopus subject areas

    • Sociology and Political Science

    Cite this

    Aronow, P. M., Samii, C., & Assenova, V. A. (2015). Cluster-robust variance estimation for dyadic data. Political Analysis, 23(4), 564-577. [mpv018]. https://doi.org/10.1093/pan/mpv018

    Cluster-robust variance estimation for dyadic data. / Aronow, Peter M.; Samii, Cyrus; Assenova, Valentina A.

    In: Political Analysis, Vol. 23, No. 4, mpv018, 01.10.2015, p. 564-577.

    Research output: Contribution to journalArticle

    Aronow, PM, Samii, C & Assenova, VA 2015, 'Cluster-robust variance estimation for dyadic data', Political Analysis, vol. 23, no. 4, mpv018, pp. 564-577. https://doi.org/10.1093/pan/mpv018
    Aronow PM, Samii C, Assenova VA. Cluster-robust variance estimation for dyadic data. Political Analysis. 2015 Oct 1;23(4):564-577. mpv018. https://doi.org/10.1093/pan/mpv018
    Aronow, Peter M. ; Samii, Cyrus ; Assenova, Valentina A. / Cluster-robust variance estimation for dyadic data. In: Political Analysis. 2015 ; Vol. 23, No. 4. pp. 564-577.
    @article{c81b5360b5e64e58ba62cf5746dd9d80,
    title = "Cluster-robust variance estimation for dyadic data",
    abstract = "Dyadic data are common in the social sciences, although inference for such settings involves accounting for a complex clustering structure. Many analyses in the social sciences fail to account for the fact that multiple dyads share a member, and that errors are thus likely correlated across these dyads. We propose a non-parametric, sandwich-type robust variance estimator for linear regression to account for such clustering in dyadic data. We enumerate conditions for estimator consistency. We also extend our results to repeated and weighted observations, including directed dyads and longitudinal data, and provide an implementation for generalized linear models such as logistic regression. We examine empirical performance with simulations and an application to interstate disputes.",
    author = "Aronow, {Peter M.} and Cyrus Samii and Assenova, {Valentina A.}",
    year = "2015",
    month = "10",
    day = "1",
    doi = "10.1093/pan/mpv018",
    language = "English (US)",
    volume = "23",
    pages = "564--577",
    journal = "Political Analysis",
    issn = "1047-1987",
    publisher = "Oxford University Press",
    number = "4",

    }

    TY - JOUR

    T1 - Cluster-robust variance estimation for dyadic data

    AU - Aronow, Peter M.

    AU - Samii, Cyrus

    AU - Assenova, Valentina A.

    PY - 2015/10/1

    Y1 - 2015/10/1

    N2 - Dyadic data are common in the social sciences, although inference for such settings involves accounting for a complex clustering structure. Many analyses in the social sciences fail to account for the fact that multiple dyads share a member, and that errors are thus likely correlated across these dyads. We propose a non-parametric, sandwich-type robust variance estimator for linear regression to account for such clustering in dyadic data. We enumerate conditions for estimator consistency. We also extend our results to repeated and weighted observations, including directed dyads and longitudinal data, and provide an implementation for generalized linear models such as logistic regression. We examine empirical performance with simulations and an application to interstate disputes.

    AB - Dyadic data are common in the social sciences, although inference for such settings involves accounting for a complex clustering structure. Many analyses in the social sciences fail to account for the fact that multiple dyads share a member, and that errors are thus likely correlated across these dyads. We propose a non-parametric, sandwich-type robust variance estimator for linear regression to account for such clustering in dyadic data. We enumerate conditions for estimator consistency. We also extend our results to repeated and weighted observations, including directed dyads and longitudinal data, and provide an implementation for generalized linear models such as logistic regression. We examine empirical performance with simulations and an application to interstate disputes.

    UR - http://www.scopus.com/inward/record.url?scp=84944403658&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84944403658&partnerID=8YFLogxK

    U2 - 10.1093/pan/mpv018

    DO - 10.1093/pan/mpv018

    M3 - Article

    VL - 23

    SP - 564

    EP - 577

    JO - Political Analysis

    JF - Political Analysis

    SN - 1047-1987

    IS - 4

    M1 - mpv018

    ER -