Bartlett’s correction in measurement errors models

Correção de Bartlett em modelos com erros nas vari´aveis

Authors

DOI:

https://doi.org/10.5902/2179460X83861

Keywords:

Bartlett correction, Bivariate normal distribution, Errors-in-Variables models, Small-Sample regression

Abstract

In this work, we address testing of simple linear regression models with errors in the variables when the sample size is small. Such models are generalizations of classical regression models and have been applied in various fields such as medicine and chemistry. The likelihood ratio statistic can be used to perform tests on the parameters of these models. However, in small samples, the chi-squared distribution does not provide a acurate approximation for the distribution of this statistic under the null hypothesis. Therefore, we construct a Bartlett correction factor for the likelihood ratio test to reduce the distortion in test size, as this modified statistic has a chi-squared distribution up to an error of order o(n−1), where n is the sample size. An advantage of the employed methodology is that the expression for Bartlett’s correction involves only simple matrix and vector operations. We present a simulation study comparing the corrected and uncorrected statistics. The numerical results favor the modified test. To illustrate the proposed methodology, we demonstrate an application using real data. Additionally, we develop a routine using the R language where it is possible to obtain the statistic for the likelihood ratio test, corrected by a Bartlett correction factor, for the studied model.

Downloads

Download data is not yet available.

Author Biographies

Kaíque Ferreira Henrique de Souza, Universidade Federal de Goiás

He holds a degree in Statistics from the Federal University of Goiás. He is currently a FullSack React and Node.

Tatiane Ferreira do Nascimento Melo, Universidade Federal de Goiás

I have a bachelor's degree in Mathematics from the Federal University of Goiás (2002), a master's degree in Statistics from the Federal University of Pernambuco (2004), a Ph.D. (2009), and a postdoctoral degree (2015) in Statistics from the University of São Paulo. Currently, I am an Associate Professor III at the Federal University of Goiás. I have experience in the field of Parametric Inference, with an emphasis on refining asymptotic methods and regression models.

Eder Angelo Milani, Universidade Federal de Goiás

I have a bachelor's degree in Applied and Computational Mathematics from São Paulo State University (2008), a master's degree in Statistics from the Federal University of São Carlos (2011), and a Ph.D. in Statistics from the Federal University of São Carlos (2016). I have experience in the field of Probability and Statistics, primarily focusing on survival analysis and time series.

Tiago Moreira Vargas, Universidade Federal de Goiás

He holds a Bachelor's and a Master's degree in Mathematics from the Federal University of Goiás and a Ph.D. in Statistics from the Institute of Mathematics and Statistics at the University of São Paulo (USP - Capital). He served as an Assistant Professor at the Faculty of Mathematics of the Federal University of Uberlândia (UFU) from 2013 to 2017. Since 2017, he has been an Assistant Professor at the Institute of Mathematics and Statistics of the Federal University of Goiás. His primary area of ​​interest is the refinement of asymptotic methods.

References

Aoki, R., Bolfarine, H., & Singer, J. M. (2001). Null intercept measurement error regression models. Test, 10(A):441–457.

Araújo, M. C., Cysneiros, A. H. M. A., & Montenegro, L. C. (2022). Bartlett and bartlett-type corrections in heteroscedastic symmetric nonlinear regression models. An Acad Bras Cienc, 21:1–15.

Aubin, E. C. Q. & Cordeiro, G. M. (2000). Bartlett-corrected tests for normal linear models when the error covariance matrix is nonscalar. Communications in Statistics – Theory and Methods, 29:2405–2426.

Barndorff-Nielsen, O. (1986). Inference on full or partial parameters, based on the standardized signed log likelihood ratio. Biometrika, 73:307–322.

Bartlett, M. S. (1937). Properties of sufficiency and statistical tests. Proceedings of the Royal Society of London A, 160:268–282.

Bartlett, M. S. (1947). Multivariate analysis. Journal of the Royal Statistical Society. Supplement, 9:179–197.

Bartlett, M. S. (1954). A note on the multiplying factors for various χ2 approximations. Journal of the Royal Statistical Society B, 16:296–298.

Bayer, F. M. & Cribari-Neto, F. (2013). Bartlett corrections in beta regression models. Journal of Statistical Planning and Inference, 143:531–547.

Chan, L. K. & Mak, T. K. (1979). On the maximum likelihood estimation of a linear structural relationship when the intercept is known. Journal of Multivariate Analysis, 9:304–313.

Cheng, C. L. & Van Ness, J. W. (1999). Statistical Regression with Measurement Error. Oxford University Press.

Cordeiro, G. M. (1993). General matrix formulae for computing bartlett corrections. Statistics and Probability Letters, 16:11–18.

Cordeiro, G. M. & Cribari-Neto, F. (1996). On bartlett and bartlett-type corrections. Econometric Reviews, 15:339–367.

Cordeiro, G. M. & Cribari-Neto, F. (2014). An Introduction to Bartlett Correction and Bias Reduction. Springer.

Fuller, W. A. (1987). Measurement Error Models. Wiley.

Kulathinal, S. B., Kuulasmaa, K., & Gasbarra, D. (2002). Estimation of an errors-invariables regression model when the variances of the measurement errors vary between the observations. Statistics in Medicine, 21:1089–1101.

Magalhães, T. M., Pereira, G. H. A., Botter, D. A., & Sandoval, M. C. (2024). Bartlett corrections for zero-adjusted generalized linear models. Statistical Papers, 65:2191–2209.

Melo, T. F. N., Vargas, T. M., Lemonte, A. J., & Moreno-Arenas, G. (2022). Higherorder asymptotic refinements in the multivariate dirichlet regression model. Communications in Statistics – Simulation and Computation, 51:53–71.

Melo, T. F. N., Vargas, T. M., Lemonte, A. J., & Patriota, A. G. (2024). Higherorder asymptotic refinements in a multivariate regression model with general parameterization. Journal of Statistical Computation and Simulation, 94(13):2952–2975.

Neyman, J. & Pearson, E. S. (1928). On the use and interpretation of certain test criteria for purposes of statistical inference. Biometrika, 20:175–240.

Noma, H., Nagashima, K., Maruo, K., Gosho, M., & Furukawa, T. A. (2018). Bartlett-type corrections and bootstrap adjustments of likelihood-based inference methods for network meta-analysis. Statistics in Medicine, 37:1178–1190.

R Development Core Team (2022). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.

Riu, J. & Rius, F. X. (1996). Assessing the accuracy of analytical methods using linear regression with errors in both axes. Analytical Chemistry, 68:1851–1857.

Wong, M. Y. (1991). Bartlett adjustment to the likelihood ratio statistic for testing several slopes. Biometrika, 78:221–224.

Wong, O., Qiao, G., Di Loreto, A., Shi, X., & Wong, A. (2025). Bartlett-type correction for testing homogeneity of inverse gaussian means. Journal of Statistical Computation and Simulation, pages 1–20.

Zavala, A. A. Z., Bolfarine, H., & De Castro, M. (2007). Assessing the accuracy of analytical methods using linear regression with errors in both axes. Annals of the Institute of Statistical Mathematics, 59:515–530.

Published

2026-09-25