Portal de Periódicos da CAPES

Estimation of the Slope and Analysis of Covariance When the Concomitant Variable is Measured with Error

1972; Volume: 67; Issue: 340 Linguagem: Inglês

10.1080/01621459.1972.10481321

1537-274X

James S. DeGracie, Wayne A. Fuller,

Statistical Methods and Bayesian Inference

Abstract Estimators of the slope for models with errors of observation in the explanatory variable are investigated. Estimators are developed which have a bias of smaller order and smaller mean square error than the common estimator. These estimators of slope are utilized in the analysis of covariance where the concomitant variable is measured with error. The procedure is applicable for any number of treatments or determinations per treatment. The computations are illustrated with an example.