Kybernetika 53 no. 1, 26-58, 2017

Instrumental weighted variables under heteroscedasticity, Part II - Numerical study

Jan Ámos VíšekDOI: 10.14736/kyb-2017-1-0026

Abstract:

Results of a numerical study of the behavior of the instrumental weighted variables estimator - in a competition with two other estimators - are presented. The study was performed under various frameworks (homoscedsticity/heteroscedasticity, several level and types of contamination of data, fulfilled/broken orthogonality condition). At the beginning the optimal values of eligible parameters of estimatros in question were empirically established. It was done under the various sizes of data sets and various levels of the contamination of data. These values were then utilized in the numerical study. Its results indicate that instrumental weighted variables are as good as $S$- and $W$-$estimators$ and under heteroscedasticity even better. The weight function of Tukey's type was used.

Keywords:

heteroscedasticity of disturbances, numerical study of instrumental weighted variables.

Classification:

62J02, 62F35

paper.pdf

References:

  1. D. F. Andrews, P. J. Bickel, F. R. Hampel, P. J. Huber, W. H. Rogers and J. W. Tukey: Robust Estimates of Location: Survey and Advances. Princeton University Press, Princeton 1972.   CrossRef
  2. J. Antoch and J. Á. Víšek: Robust estimation in linear models and its computational aspects. In: Contributions to Statistics: Computational Aspects of Model Choice (J. Antoch, ed.), Springer Verlag 1992, pp. 39-104.   DOI:10.1007/978-3-642-99766-2_4
  3. A. C. Atkinson and M. Riani: Robust Diagnostic Regresion Analysis    CrossRef
  4. O. Barndorff-Nielsen and D. R. Cox: Edgeworth and saddle-point approximations with statistical applications. J. Royal Statist. Soc. B 41 (1979), 3, 279-312.   CrossRef
  5. V. Benáček and J. Á. Víšek: Determining factors of trade specialization and growth of a small economy in transition. Impact of the EU opening-up on Czech exports and imports. IIASA, IR series, no. IR-03-001, 2002, pp. 1-41.   CrossRef
  6. P. J. Bickel: One-step Huber estimates in the linear model. JASA 70 (1975), 428-433.   DOI:10.2307/2285834
  7. N. A. Campbell, H. P. Lopuhaa and P. J. Rousseeuw: On calculation of a robust $S$-estimator if a covariance matrix. Statist. Medcine 17 (1998), 2685-2695.   DOI:10.1002/(sici)1097-0258(19981215)17:23<2685::aid-sim35>3.3.co;2-n
  8. S. Chatterjee and A. S. Hadi: Sensitivity Analysis in Linear Regression. J. Wiley and Sons, New York 1988.   DOI:10.1002/9780470316764
  9. P. Čížek: Generalized Method of Trimmed Moments. Center Discussion Paper 2009-25, Tilburg University 2009.   DOI:10.2139/ssrn.1382807
  10. G. V. Cohen-Freue, H. Ortiz-Molina and R. H. Zamar: Natural robustification of the ordinary instrumental variables estimator. Biometrics 69 (2013), 641-650.   DOI:10.1111/biom.12043
  11. R. Desborges and V. Verardi: A robust instrumental-variable estimator. The Stata J. 12 (2012), 169-181.   CrossRef
  12. F. Y. Edgeworth: On observations relating to several quantities. Phil. Mag. (5th Series) 24 (1887), 222.   CrossRef
  13. C. A. Field and E. M. Ronchetti: Small Sample Asymptotics. Institute of Mathematical Statistics Monograph Series, Hayward 1990.   CrossRef
  14. D. Gervini V. J. and Yohai: A class of robust and fully efficient regression estimators. Annal Statist. 30 (2002), 583-616.   DOI:10.1214/aos/1021379866
  15. P. R. Halmos: Applied mathematics is a bad mathematics. In: Mathematics Tomorrow (L. Steen, ed.), Springer Verlag, New York 1981, pp. 9-20.   DOI:10.1007/978-1-4613-8127-3_2
  16. J. Jurečková: Robust estimation of location and regression parameters and their second order asymptotic relations. In: 9th Prague Conf. on Inform. Theory, Reidel 1983, pp. 19-32.   DOI:10.1007/978-94-009-7013-7_2
  17. J. Jurečková and P. K. Sen: Uniform second order asymptotic linearity of $M$-statistics in linear models. Statistics and Decisions 7 (1989), 263-276.   DOI:10.1524/strm.1989.7.3.263
  18. J. Jurečková and M. Malý: The asymptotics for studentized $k$-step $M$-estimators of location. Sequential Analysis 14 (1995), 229-245.   DOI:10.1080/07474949508836334
  19. J. Jurečková and A. H. Welsh: Asymptotic relations between $L$- and $M$-estimators in the linear model. Ann. Inst. Statist. Math. 42 (1990), 671-698.   DOI:10.1007/bf02481144
  20. R. Koenker and G. Bassett: Regression quantiles. Econometrica 46 (1978), 33-50.   DOI:10.2307/1913643
  21. R. A. Maronna and V. J. Yohai: Asymptotic behaviour of general $M$-estimates for regression and scale with random carriers. Z. Wahrscheinlichkeitstheorie verw. Gebiete 58 (1981), 7-20.   DOI:10.1007/bf00536192
  22. L. Mašíček: Optimality of the least weighted squares estimator. Kybernetika 40 (2004), 715-734.   CrossRef
  23. J. Robinson: Saddle point approximations for permutation tests and confidence intervals. J. Roy. Statist. Soc. B 44 (1982), 1, 91-101.   CrossRef
  24. P. J. Rousseeuw: Multivariate estimation with high breakdown point. In: Proc. Fourth Pannonian Symposium on Mathematical Statistics, Bad Tatzmannsdorf 1983, pp. 283-297.   DOI:10.1007/978-94-009-5438-0_20
  25. P. J. Rousseeuw: Least median of square regression. J. Amer. Statist. Association 79 (1984), 871-880.   DOI:10.1080/01621459.1984.10477105
  26. P. J. Rousseeuw and A. M. Leroy: Robust Regression and Outlier Detection. J. Wiley and Sons, New York 1987.   DOI:10.1002/0471725382
  27. P. J. Rousseeuw and V. Yohai: Robust regressiom by means of $S$-estimators. In: Robust and Nonlinear Time Series Analysis (J. Franke, W. H\H{a}rdle, and R. D. Martin, eds.), Lecture Notes in Statistics No. 26, Springer Verlag, New York 1984, pp. 256-272.   DOI:10.1007/978-1-4615-7821-5_15
  28. A. M. Rubio and J. Á. Víšek: A note on asymptotic linearity of $M$-statistics in non-linear models. Kybernetika 32 (1996), 353-374   CrossRef
  29. A. F. Siegel: Robust regression using repeated medians. Biometrica 69 (1982), 242-244.   DOI:10.1093/biomet/69.1.242
  30. V. Verardi and A. McCathie: The $S$-estimator of multivariate location and scatter in Stata. The Stata J. 12 (2012), 299-307.   CrossRef
  31. A. H. Welsh: Bahadur representation for robust scale estimators based on regression residuals. Ann. Statist. 14 (1986), 1246-1251.   DOI:10.1214/aos/1176350064
  32. J. Á. Víšek: On second order efficiency of a robust test and approximations of its error probabilities. Kybernetika 8 (1983), 3, 387-407.   CrossRef
  33. J. Á. Víšek: Estimating contamination level. In: Proc. Fifth Pannonian Symposium on Mathematical Statistics, Visegrád 1985, pp. 401-414.   CrossRef
  34. J. Á. Víšek: Robust instruments. In: Robust'98 (J. Antoch and G. Dohnal, eds.), Union of Czechoslovak Mathematicians and Physicists 1998, pp. 195-224.   CrossRef
  35. J. Á. Víšek: The robust regression and the experiences from its application on estimation of parameters in a dual economy. In: Proc. Conference Macromodels'99, Wroclaw University 1999, pp. 424-445.   CrossRef
  36. J. Á. Víšek: Regression with high breakdown point. In: ROBUST 2000 (J. Antoch and G. Dohnal, eds.), The Union of the Czech Mathematicians and Physicists and the Czech Statistical Society 2001, pp. 324-356.   CrossRef
  37. J. Á. Víšek: Character of the Czech economy in transition. In: Proc. The Czech society on the break of the third millennium, Karolinum, Publishing House of the Charles University 2000, pp. 181-205.   CrossRef
  38. J. Á. Víšek: The least trimmed squares. Part I - Consistency. Kybernetika 42 (2006), 1-36.   CrossRef
  39. J. Á. Víšek: Robust error-term-scale estimate. IMS Coll., Nonparametrics and Robustness in Modern Statistical Inference and Time Series Analysis: Festschrift for Jana Jurečková 7 (2010), 254-267.   CrossRef
  40. J. Á. Víšek: Least weighted squares with constraints and under heteroscedasticity. Bull. Czech Econometr. Soc. 20 (2012), 31, 21-54.   CrossRef
  41. J. M. Wooldridge: Introductory Econometrics. A Modern Approach. MIT Press, Cambridge 2006. Second edition 2009.   CrossRef