Robust Statistics applied to location and scale measures: Technical Note.

Keywords: robust statistics, robust location measures, robust scale measures, outliers, skewed distributions

Abstract

When measurement of a variable is conducted, it is common for estimates of a sample to have a substantial amount of bias. Some factors responsible of this outcome are asymmetry and outliers. In psychology and the social sciences, it is usual to find that most of the commonly used statistics, such as the arithmetic mean and its associated standard error are, at best, imprecise in the task of data summary and inference. To overcome this situation, researchers can make use of robust statistics. Robust statistics provide a series of alternative estimates resistant to the influence of outliers, resulting in more precise information and inferences. The aim of this paper is to describe a group of procedures to calculate measures of location and scale with robust methods using R and IBM SPSS software. First, the different visual and quantitative methods for detecting outliers are reviewed. Then, different alternatives of location measures are reviewed such as the trimmed mean, the winsorized mean and M estimators. Each location measure will be presented with its associated standard error. Last, some scale measures are presented, such as the interquartile range and a modification, called ideal fourths. Conclusions emphasize the thoughtful use of the procedures in relation to the reader possibilities, interests and the theoretical and methodological implications.

Downloads

Download data is not yet available.

Author Biographies

David Ruiz-Méndez, | Instituto de Estudios Superiores Monterrey |
Licenciado y maestro en psicología por la UNAM. Ha participado en Congresos especializados internacionales en relación con la evaluación psicológica en las organizaciones y análisis experimental y cuantitativo en psicología. Actualmente estudia el doctorado en Análisis Experimental de la Conducta en la UNAM y se desempeña como profesor en el Tecnológico de Monterrey Campus Estado de México. 
Mirna Elizabeth Quezada, | Universidad Nacional Autónoma de México | Facultad de Estudios Superiores Iztacala |
Licenciada en Psicóloga por Universidad Autónoma de Tamaulipas; Maestra en Psicología, Gestión Organizacional por la Facultad de Estudios Superiores Iztacala UNAM, becaria de los programas PAPIME, PAPIIT y CONACyT. Actual docente del programa de Maestría en Psicología de la UNAM, de la Lic. en Psicología del Sistema de Universidad Abierta y Educación a Distancia-UNAM, y docente de la Licenciatura en línea de la Universidad Tecnológica de México. 
Cynthia Zaira Vega Valero, | Universidad Nacional Autónoma de México | Facultad de Estudios Superiores Iztacala |
Doctora en Psicología por la UNAM, profesora titular de la carrera de Psicología y coordinadora de la residencia en Gestión Organizacional de la maestría en Psicología. de la FES Iztacala, Universidad Nacional Autónoma de México. 

References

Casella, G. & Berger, R. (2002). Statistical Inference (2 ed.). Pacific Grove: Duxbury.

Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences. 2nd edition. Academic Press, New York.

Emerson, J. D. & Strenio, J. (1983). Boxplots and Batch comparison. En D. C. Hoaglin, F. Mosteller, & J.W. Tukey (Eds.), Understanding Robust and Exploratory Data Analysis. New York: Wiley.

Fernández, S. F., Sánchez, J. M. C., Córdoba, A., & Largo, A. C. (2002). Estadística descriptiva. España: Esic Editorial.

Field, A. (2013). Discovering statistics using IBM SPSS statistics. London: Sage.

Field A. & Wilcox R. (2017). “Robust statistical methods: A primer for clinical psychology and experimental psychopathology researchers.” Behaviour Research and Therapy, 98, 19 38.

Goodall, C. (1983). M-Estimators of location: An outline of the theory. En D. C. Hoaglin, F. Mosteller, & J.W. Tukey (Eds.), Understanding Robust and Exploratory Data Analysis. New York: Wiley.

Hopkins, K. & Glass, G. (1978) Basic Statistics for the Behavioral Sciences. Englewood Cliffs, New Jersey: Prentice-Hall.

Huber, P. (1981) Robust statistics, New York: John Wiley & Sons, Inc.

Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill, Inc.

Press, W., Teukolsky, S., Vetterling, W. & Flannery, B (2007). Numerical Recipes 3rd Edition: The Art of Scientific Computing (3rd. ed.). Cambridge University Press, USA.

Rosenberg, J. L. & Gasko, M. (1983). Comparing Location Estimators: Trimmed Means, Medians, and Trimean. En D. C. Hoaglin, F. Mosteller, & J.W. Tukey (Eds.), Understanding Robust and Exploratory Data Analysis. New York: Wiley.

Rousseeuw, P. J. (1991). Tutorial to robust statistics. Journal of Chemometrics, 5(1), 1-20.

Ruiz, D. (2020). Distribución de la conducta operante en humanos: desarrollo de preferencia y parámetros de reforzamiento relevantes. Tesis doctoral no publicada. Universidad Nacional Autónoma de México.

Tukey, J. W. (1960). A survey of sampling from contaminated normal distributions. In I. Olkin, S. Ghurye, W. Hoeffding, W. Madow, & H. Mann (Eds.), Contributions to Probability and Statistics. Stanford, CA: Stanford University Press.

Wilcox, R. (2017). Introduction to Robust Estimation & Hypothesis Testing. 4th edition. Elsevier, Amsterdam, The Netherlands.

Published
2020-07-11
How to Cite
Ruiz-Méndez, D., Quezada, M. E., & Vega Valero, C. Z. (2020). Robust Statistics applied to location and scale measures: Technical Note. Revista Digital Internacional De Psicología Y Ciencia Social, 6(2), 499-517. https://doi.org/10.22402/j.rdipycs.unam.6.2.2020.302.499-517