eduzhai > Physical Sciences > Physics Sciences >

Longitudinal Survey, Nonmonotone, Nonresponse, Imputation, Nonparametric Regression

  • Save

... pages left unread,continue reading

Document pages: 17 pages

Abstract: The study focuses on theimputation for the longitudinal survey data which often has nonignorablenonrespondents. Local linear regression is used to impute the missing valuesand then the estimation of the time-dependent finite populations means. The asymptoticproperties (unbiasedness and consistency) of the proposed estimator areinvestigated. Comparisons between different parametric and nonparametricestimators are performed based on the bootstrap standard deviation, mean squareerror and percentage relative bias. A simulation study is carried out todetermine the best performing estimator of the time-dependent finite populationmeans. The simulation results show that local linear regression estimatoryields good properties.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×