Multiple Imputation for Missing Network Data

Robert Krause

    Research output: ThesisThesis fully internal (DIV)

    242 Downloads (Pure)

    Abstract

    In this thesis we developed, implemented, and evaluated multiple imputation algorithms for missing network data. The algorithms are able to handle cross-sectional, longitudinal,and multiplex network structures, as well as nodal attributes (coevolving behaviors). They were implemented for the two most important statistical network model families in the social sciences, that is, Exponential Random Graph Models and Stochastic Actor-oriented Models.
    Original languageEnglish
    QualificationDoctor of Philosophy
    Awarding Institution
    • University of Groningen
    Supervisors/Advisors
    • Snijders, Thomas, Supervisor
    • Huisman, Mark, Co-supervisor
    • Steglich, Christian, Co-supervisor
    • Albers, Casper, Assessment committee
    • van Buuren, S, Assessment committee, External person
    • Veenstra, René, Assessment committee
    Award date19-Dec-2019
    Place of Publication[Groningen]
    Publisher
    Print ISBNs978-94-034-1984-8
    Electronic ISBNs978-94-034-1983-1
    DOIs
    Publication statusPublished - 2019

    Cite this