Publication: Estadísticos para la detección de observaciones anómalas en modelos de elección binaria: una aplicación con datos reales
Loading...
Files
Official URL
Full text at PDC
Publication Date
1994
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Facultad de Ciencias Económicas y Empresariales. Instituto Complutense de Análisis Económico (ICAE)
Citations
Exportar
Abstract
Este trabajo trata el problema de la detección de observaciones anómalas en modelos de elección binaria. Partiendo del estadístico propuesto en Gracia-Díez y Serrano (1994) que mide la influencia individual de cada observación sobre el vector de parámetros estimado, se derivan otros estadísticos que evalúan la influencia individual y de grupos de observaciones sobre i) el vector de probabilidades estimadas e ii) sobre subconjuntos de parámetros y combinaciones lineales de los mismos. También, se generaliza el método de Peña y Yohai (1991) para la detección de observaciones enmascaradas en modelos lineales al caso de los modelos de elección binaria. Finalmente, se propone una estrategia de diagnosis para la detección de anomalías en este tipo de modelos. Esta estrategia se ilustra mediante su aplicación al modelo probit estimado por Dhillon el. al (1987).
This paper considers the problem of outliers in binary response models. Based on the statistic proposed by Gracia-Díez y Serrano (1994) which measures the influence of each observation on the estimated parameler vector, we derive other statistics in order to measure the influence of each observation as well as the influence of a group of observations on i) the vector of estimated probabilities and ii) subsets and linear combinations of tbe parameters in the model. Also, the method proposed by Peña y Yohai (1991) to deal with the masking problem in linear models has been generalised to the case of binary choice models. Lastly, we propose a diagnostic strategy to detect outliers in this type of models. The application of this strategy is ilustrated by estimating the probit model used by Dhillon et. al (1987).
This paper considers the problem of outliers in binary response models. Based on the statistic proposed by Gracia-Díez y Serrano (1994) which measures the influence of each observation on the estimated parameler vector, we derive other statistics in order to measure the influence of each observation as well as the influence of a group of observations on i) the vector of estimated probabilities and ii) subsets and linear combinations of tbe parameters in the model. Also, the method proposed by Peña y Yohai (1991) to deal with the masking problem in linear models has been generalised to the case of binary choice models. Lastly, we propose a diagnostic strategy to detect outliers in this type of models. The application of this strategy is ilustrated by estimating the probit model used by Dhillon et. al (1987).
Description
UCM subjects
Unesco subjects
Keywords
Citation
Amemiya, T. (1981). "Qualilative Response Models: A Survey". Journal of Economic Literature, XIX, 1483-1536.
Amemiya, T. (1985). Advanced Econometrics, Oxford, Basil Blackwell Lid.
Atkinson, A. C. (1985). Plots, Transformations and Regression, New York, Oxford University Press.
Bedrick. E. J. Y J. R. Hill (1990). "Outlier Tests for Logistic Regression, a Conditional Approach", Biometrika, 77, 4, 815-821.
Belsley, D. A., E. Kuh Y R. E. Welsch (1981). Regression Diagnostics. Identifying influencial Data and Sources of Collinearity, New York, John Wiley & Sons.
Cook, R. D. (1977). "Defection of Influential Observation in Linear Regression". Technometrics, 19, 1, 15-18.
Copas, J. B. (1988). "Binary Regression Models for Contaminated Data", Joumal of the Royal Statistical Society, B, 50, 2, 225-265.
Dhillon, U. S., J. D. Shilling y C. F. Sirmans (1981). "Choosing between fixed and Adjustable Rate Mortgages". Journal of Money, Credit and Banking, 19, 1, 260-267.
Gracia-Diez, M. y G. R. Serrano (1994), "Observaciones Anómalas en Modelos de Elección Binaria". Estadística Española. próxima aparición.
Jennings, D. E. (1986). "Outliers and Residual Distributions in Logistic Regression". Journal of the American Statistical Association, 81, 396. 987-990.
Lott, W. F. y S. C. Kay (1992). Applied Econometrics: Problems with Data Sets, The Dryden Press.
McCullagh, P. y J. A. Nelder (1983). Generalized Linear Models, London: Chapman and Hall, Inc.
Peña, D. (1987). "Observaciones Influyentes en Modelos Econométricos". Investigaciones Económicas, XI, 1, 3-24.
Peña, D. y V. J. Yohai (1991). "The Detection of influential Subsets in Linear Regression using an influence Matrix". mimeo.
Pregibon, D. (1981). "Logistic Regression Diagnostics", The Annals of Statistics, 9, 4, 705-724.
Ralston, A. y P. Rabinowitz (1978). A First Course in Numerical Analysis. McGraw-Hill.
Rousseeuw, P. J. y B. C. van Zomeren (1990). "Unmasking Multivariate Outliers and Leverage Points", con discusión. Journal of de American Statistical Association, 85, 411, 633-651.
Serrano, G. R. (t993). Observaciones anómalas en modelos de variable dependiente cualitativa. Tesis doctoral. Universidad Complutense de Madrid.
Smith, B. T., J. M. Boyle, J. J. Dongarra, B. S. Garbow, Y. Ikebe, V. C. Klema y C. B. Moler (1974). Matrix Eigensystem Routines. EISPACK Guide, Springer-Verlag.
Williams, D. A. (1987). "Generalized linear Model Diagoostics: The Deviance and Single Case Deletion", Applied Statistics, 36, 2, 181-191.