Principal components estimator for measurement error models

ABSTRACT

In this paper, we carry out the principal components regression approach to the measurement error models. We introduce the principal components estimator and then the restricted principal components estimator by combining the approaches principal components regression estimator and restricted least squares estimator for the measurement error models, when the reliability matrix known and unknown, separately. We investigate the asymptotic properties and matrix mean squared error performances of the new estimators. Also, we conduct a Monte Carlo simulation study and a numerical example to investigate the performances of the proposed estimators by the scalar mean squared error criterion.

KEYWORDS:

• Measurement error
• multicollinearity
• reliability matrix
• principal components estimator
• restricted estimator

AMS classification:

• 62J05

Acknowledgements

The author is most grateful to Prof. Dr Abdol Rahman Rasekh from Shahid Chamran University Ahvaz for providing Egyptian Pottery dataset.

Disclosure statement

No potential conflict of interest was reported by the author.