Modelling the convergence probability of a confidence interval on the among-group variance component in the unbalanced random one-way model (Q2488400)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Modelling the convergence probability of a confidence interval on the among-group variance component in the unbalanced random one-way model
scientific article

    Statements

    Modelling the convergence probability of a confidence interval on the among-group variance component in the unbalanced random one-way model (English)
    0 references
    0 references
    0 references
    0 references
    24 May 2006
    0 references
    The one-way Gaussian ANOVA model \(y_{ij}=\mu+\alpha_i+\varepsilon_{ij}\), \(i=1\),\dots,\(k\), \(j=1\),\dots,\(n_k\), is considered where \(\alpha_i\) and \(\varepsilon_{ij}\) are independent Gaussian random variables with zero mean and variances \(\sigma_\alpha^2\) and \(\sigma_\varepsilon^2\), respectively. Four procedures are considered for deriving approximate confidence intervals on \(\sigma_\alpha^2\): a modified large sample procedure, the Thomas-Hultquist procedure, a modified harmonic mean procedure, and the Burdick-Eickman procedure. The dependence of the true coverage probability on \(k\), \(n=\sum_{i=1}^k n_i\), \(\phi=n^2/(k\sum_{i=1}^k n_i^2)\) and \(\rho=\sigma_\alpha^2/(\sigma_\alpha^2+\sigma_\varepsilon^2)\), is estimated by simulations using a generalized linear model (GLM) fit. Resulting parameters of GLM models for all four procedures are given in tables.
    0 references
    Thomas-Hultquist confidence interval
    0 references
    harmonic mean
    0 references
    Burdick-Eickman confidence interval
    0 references
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references