Conservative variance estimation for sampling designs with zero pairwise inclusion probabilities

Peter M. Aronow, Cyrus Samii

    Research output: Contribution to journalArticle

    Abstract

    We consider conservative variance estimation for the Horvitz-Thompson estimator of a population total in sampling designs with zero pairwise inclusion probabilities, known as "non-measurable" designs. We decompose the standard Horvitz-Thompson variance estimator under such designs and characterize the bias precisely. We develop a bias correction that is guaranteed to be weakly conservative (nonnegatively biased) regardless of the nature of the non-measurability. The analysis sheds light on conditions under which the standard Horvitz-Thompson variance estimator performs well despite non-measurability and where the conservative bias correction may outperform commonly-used approximations.

    Original languageEnglish (US)
    Pages (from-to)231-241
    Number of pages11
    JournalSurvey Methodology
    Volume39
    Issue number1
    StatePublished - 2012

    Fingerprint

    Horvitz-Thompson Estimator
    Inclusion Probabilities
    Variance Estimation
    Sampling Design
    Pairwise
    Bias Correction
    Variance Estimator
    Sampling
    Zero
    Biased
    Decompose
    Approximation
    Standards
    Design

    Keywords

    • Horvitz-Thompson estimation
    • Non-measurable designs
    • Variance estimation

    ASJC Scopus subject areas

    • Modeling and Simulation
    • Statistics and Probability

    Cite this

    Conservative variance estimation for sampling designs with zero pairwise inclusion probabilities. / Aronow, Peter M.; Samii, Cyrus.

    In: Survey Methodology, Vol. 39, No. 1, 2012, p. 231-241.

    Research output: Contribution to journalArticle

    Aronow, Peter M. ; Samii, Cyrus. / Conservative variance estimation for sampling designs with zero pairwise inclusion probabilities. In: Survey Methodology. 2012 ; Vol. 39, No. 1. pp. 231-241.
    @article{e18fd6b6975a400fb1fe101f78b097dd,
    title = "Conservative variance estimation for sampling designs with zero pairwise inclusion probabilities",
    abstract = "We consider conservative variance estimation for the Horvitz-Thompson estimator of a population total in sampling designs with zero pairwise inclusion probabilities, known as {"}non-measurable{"} designs. We decompose the standard Horvitz-Thompson variance estimator under such designs and characterize the bias precisely. We develop a bias correction that is guaranteed to be weakly conservative (nonnegatively biased) regardless of the nature of the non-measurability. The analysis sheds light on conditions under which the standard Horvitz-Thompson variance estimator performs well despite non-measurability and where the conservative bias correction may outperform commonly-used approximations.",
    keywords = "Horvitz-Thompson estimation, Non-measurable designs, Variance estimation",
    author = "Aronow, {Peter M.} and Cyrus Samii",
    year = "2012",
    language = "English (US)",
    volume = "39",
    pages = "231--241",
    journal = "Survey Methodology",
    issn = "0714-0045",
    publisher = "Statistics Canada",
    number = "1",

    }

    TY - JOUR

    T1 - Conservative variance estimation for sampling designs with zero pairwise inclusion probabilities

    AU - Aronow, Peter M.

    AU - Samii, Cyrus

    PY - 2012

    Y1 - 2012

    N2 - We consider conservative variance estimation for the Horvitz-Thompson estimator of a population total in sampling designs with zero pairwise inclusion probabilities, known as "non-measurable" designs. We decompose the standard Horvitz-Thompson variance estimator under such designs and characterize the bias precisely. We develop a bias correction that is guaranteed to be weakly conservative (nonnegatively biased) regardless of the nature of the non-measurability. The analysis sheds light on conditions under which the standard Horvitz-Thompson variance estimator performs well despite non-measurability and where the conservative bias correction may outperform commonly-used approximations.

    AB - We consider conservative variance estimation for the Horvitz-Thompson estimator of a population total in sampling designs with zero pairwise inclusion probabilities, known as "non-measurable" designs. We decompose the standard Horvitz-Thompson variance estimator under such designs and characterize the bias precisely. We develop a bias correction that is guaranteed to be weakly conservative (nonnegatively biased) regardless of the nature of the non-measurability. The analysis sheds light on conditions under which the standard Horvitz-Thompson variance estimator performs well despite non-measurability and where the conservative bias correction may outperform commonly-used approximations.

    KW - Horvitz-Thompson estimation

    KW - Non-measurable designs

    KW - Variance estimation

    UR - http://www.scopus.com/inward/record.url?scp=84883034774&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84883034774&partnerID=8YFLogxK

    M3 - Article

    VL - 39

    SP - 231

    EP - 241

    JO - Survey Methodology

    JF - Survey Methodology

    SN - 0714-0045

    IS - 1

    ER -