SciELO - Scientific Electronic Library Online

vol.41 issue2Construction of the Design Matrix for Generalized Linear Mixed-Effects Models in the Context of Clinical Trials of Treatment SequencesOn Reliability in a Multicomponent Stress-Strength Model with Power Lindley Distribution author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



Related links

  • On index processCited by Google
  • Have no similar articlesSimilars in SciELO
  • On index processSimilars in Google


Revista Colombiana de Estadística

Print version ISSN 0120-1751

Rev.Colomb.Estad. vol.41 no.2 Bogotá July/Dec. 2018 

Artículos originales de investigación

Kernel Function in Local Linear Peters-Belson Regression

Función del núcleo en la regresión lineal local de Peters-Belson

Mohammad Bolbolian Ghalibaf1  a 

1 Department of Statistics, Faculty of Mathematics and Computer Science, Hakim Sabzevari Univercity, Sabzevar, Iran.


Determining the extent of a disparity, if any, between groups of people, for example, race or gender, is of interest in many fields, including public health for medical treatment and prevention of disease or in discrimination cases concerning equal pay to estimate the pay disparities between minority and majority employees. The Peters-Belson (PB) regression is a form of statistical matching, akin in spirit to Bhattacharya's bandwidth matching which is proposed for this purpose. In this paper, we review the use of PB regression in legal cases from Bura, Gastwirth & Hikawa (2012). Parametric and nonparametric approaches to PB regression are described and we show that in nonparametric PB regression a suitable kernel function can improve results, i.e. by selecting the appropriate kernel function, we can reduce bias and variance of estimators, also increase the power of tests.

Key words: Kernel Function; Local Linear Regression; Welch's Approximation


Determinar el alcance de una disparidad, si la hubiere, entre grupos de personas, por ejemplo, raza o género, es de interés en muchos campos, incluida la salud pública para el tratamiento médico y la prevención de enfermedades o en casos de discriminación en relación con la igualdad salarial para estimar las disparidades salariales entre los empleados minoritarios y mayoritarios. La regresión de Peters Belson (PB) es una forma de coincidencia estadística, similar en espíritu a la coincidencia de ancho de banda de Bhattacharya que se propone para este propósito. En este trabajo, repasamos el uso de la regresión del PB en casos legales de Bura et al. (2012). Se describen los enfoques paramétricos y no paramétricos de la regresión del PB y demostramos que en la regresión no paramétrica del PB una función de kernel adecuada puede mejorar los resultados, es decir, seleccionando la función de kernel apropiada, podemos reducir el sesgo y la varianza de los estimadores, también aumentan el poder de las pruebas.

Palabras-clave: aproximación de Welch; función kernel; regresión lineal local

Full text available only in PDF format


Achcar, J. A. & Lopes, S. R. C. (2016), `Linear and Non-Linear Regression Models Assuming a Stable Distribution', Revista Colombiana de Estadística 39(1), 109-128. [ Links ]

Belson, W. A. (1956), `A technique for studying the effects of a television broadcast', Applied Statistics 5, 195-202. [ Links ]

Bhattacharya, P. K. (1989), Estimation of treatment main effect and treatmentcovariate interaction in observational studies using bandwidth-matching, Technical Report 188, Division of Statistics, University of California, Davis. [ Links ]

Bhattacharya, P. K. & Gastwirth, J. L. (1999), `Estimation of the Odds-Ratio in an Observational Study Using Bandwidth-Matching', Journal of Nonparametric Statistics 11, 1-12. [ Links ]

Blinder, A. S. (1973), `Wage discrimination: reduced form and structural estimates', Journal of Human Resources 8, 436-455. [ Links ]

Bura, E., Gastwirth, J. L. & Hikawa, H. (2012), The use of petersbelson regression in legal cases, in `Nonparametric Statistical Methods and Related Topics: A Festschrift in Honor of Professor PK Bhattacharya on the Occasion of His 80th Birthday', World Scientific, pp. 213-231. [ Links ]

Cleveland, W. S. & Devlin, S. J. (1988), `Locally-weighted regression: an approach to regression analysis by local fitting', Journal of the American Statistical Association 83, 597-610. [ Links ]

Fan, J. & Gijbels, I. (1996), Local Polynomial Modelling and Its Applications, Chapman and Hall, London. [ Links ]

Gastwirth, J. L. (1989), `A clarification of some statistical issues in Watson vs. Fort Worth Bank and Trust', Jurimetrics Journal 29, 267-284. [ Links ]

Gastwirth, J. L. & Greenhouse, S. W. (1995), `Biostatistical Concepts and Methods in the Legal Setting', Statistics in Medicine 14, 1641-1653. [ Links ]

Gray, M. W. (1993), `Can statistics tell us what we do not want to hear? The case of complex salary structures', Statistical Science 8(2), 144-158. [ Links ]

Greiner, D. J. (2008), `Causal inference in civil rights litigation', Harvard Law Review 122, 533-598. [ Links ]

Hikawa, H. (2009), Local linear peters-belson regression and its applications to employment discrimination cases, Doctoral dissertation, Department of Statistics, George Washington University. [ Links ]

Hikawa, H., Bura, E. & Gastwirth, J. L. (2010a), `Local Linear Logistic Peters Belson Regression and its application in employment discrimination cases', Statistics and its Interface 3, 125-144. [ Links ]

Hikawa, H., Bura, E. & Gastwirth, J. L. (2010b), Robust peters-belson type estimators of measures of disparity and their applications in employment discrimination cases, Technical report, Department of Statistics, George Washington University. [ Links ]

Loader, C. (1999), Local Regression and Likelihood, Springer-Verlag, New York. [ Links ]

Nayak, T. K. & Gastwirth, J. L. (1997), `The Peters-Belson approach to measures of economic and legal discrimination', Advances in the Theory and Practice of Statistics pp. 587-601. [ Links ]

Oaxaca, R. (1973), `Male-Female differentials in urban labor markets', International Economic Review 14, 693-709. [ Links ]

Peters, C. C. (1941), `A method of matching groups for experiments with no loss of populations', Journal of Educational Research 34(8), 606-612. [ Links ]

Scheffe, H. (1970), `Practical solutions of the Behrens-Fisher problem', Journal of the American Statistical Association 65(332), 1501-1508. [ Links ]

Welch, B. L. (1949), `Further note on Mrs. Aspin's tables and on certain approximations to the tabled function', Biometrika 36(3/4), 293-296 [ Links ]

a E-mail: ;

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License