Correcting regression dilution problem using direct adjustment method

Steve Su

    Research output: Contribution to journalArticle

    Abstract

    This article presents a direct adjustment correction method to multiple regression dilution problem in the context of uncorrelated, zero mean, additive measurement errors with finite standard deviations. The proposed method gives accurate adjusted regression coefficients and standard errors to allow correct statistical inferences and outperforms the SIMEX algorithm in various simulation studies. The direct adjustment correction method is also more precise in its regression coefficient estimates compared to existing methods such as product moment correlation coefficient, intraclass coefficient, Rosner's regression and maximum likelihood estimation.
    Original languageEnglish
    Pages (from-to)25-37
    JournalFar East Journal of Theoretical Statistics
    Volume29
    Issue number1
    Publication statusPublished - 2009

    Fingerprint Dive into the research topics of 'Correcting regression dilution problem using direct adjustment method'. Together they form a unique fingerprint.

  • Cite this