A modified test for detecting influential decision-making units in data envelopment analysis

Abstract

In data analyses based on a deterministic or stochastic approach, using pre-study is very important to identify observations that are not suitable to data in general. Among such observations, those that have a high tendency to change results negatively are called influential observations. In this paper, we propose a new method to identify influential observations in Data Envelopment Analysis (DEA). Our method is a modified version of the one proposed by Pastor et al. [12]. Both methods are compared by using two well-known data sets and the outcomes are discussed. A comparative analysis indicates that our method is an effective alternative to the Pastor et al. [12] method to identify influential observations in DEA.

Keywords

• Data Envelopment Analysis,inuential observations,super efficiency,likelihood ratio test,statistical hypothesis test

References

1. Andrews, D.F. & Pregibon, D. Finding outliers that matter, J. Roy. Statist. Soc., Ser. B. 40, 85-93, 1978.
2. Bal, H., Orkcu, H.H. & Çelebioglu, S. Improving the discrimination power and weights dispersion in the data envelopment analysis, Computers & Operations Research 37, 99-107, 2010.
3. Banker, R.D., Charnes, A. & Cooper, W.W. Some models for estimating technical and scale ineciencies in data envelopment analysis, Management Science 30(9), 1078-1092, 1984.
4. Belsley, D.A., Kuh, E. & Welsch, R.E. Regression diagnostics: identifying inuential data and sources of collinearity, Willey Series in Probability and Mathematical Statistics, New York., 1980.
5. Charnes, A., Cooper, W.W. & Rhodes, E. Measuring the eciency of decision making units, European Journal of Operational Research 2(6), 429-444, 1978.
6. Charnes, A., Cooper, W.W. & Rhodes, E. Evaluating program and managerial eciency: An application of data envelopment analysis to program follow through, Management Science 27(6), 668-697, 1981.
7. Chatterjee S. & Hadi A.S. Regression analysis by example, Willey Series in Probability and Mathematical Statistics, New Jersey., 2006.
8. Cook, R.D. Detection of inuential observations in linear regression, Technometrics 19(1), 15-18, 1977.

9. Cook, R.D. & Weisberg, S.Residuals and inuence in regression, Chapman and Hall, New York. , 1982.
10. Jahanshahloo, G.R., Hosseinzadeh, F., Shoja, N., Tohidi, G. & Razavyan, S. A method for detecting inuential observations in radial DEA models, Applied Mathematics and Computation 147, 415-421, 2004.
11. Liu, J.S., Lu, Y.Y., Lu, W.M. & Lin, B.J.Y. Data Envelopment Analysis 1978-2010: A citation-based literature survey, Omega 41, 3-15, 2013.
12. Pastor, J.T., Ruiz, J.L. & Sirvent, I. A statistical test for detecting inuential observation in DEA, European Journal of Operational Research 115, 542-545, 1999.
13. Ruiz, J.L. & Sirvent, I. Techniques for the assessment of inuence in DEA, European Journal of Operational Research 132, 390-399, 2001.
14. Song, M.L., Wang, S.H. & Liu, W. A two-stage DEA approach for environmental eciency measurement, Environmental Monitoring and Assessment, 186, 3041-3051, (2014).
15. Wilson, P.W. Detecting outliers in deterministic nonparametric frontier models with mul- tiple outputs, Journal of Business and Economic Statistics 11, 319-323, 1993.
16. Wilson, P.W. Detecting inuential observations in data envelopment analysis, Journal of Productivity Analysis 6, 27-45, 1995.
17. Witte, K.D. & Marques, R.C. Inuential observations in frontier models, a robust non- oriented approach to the water sector, Annals of Operations Research 181, 377-392, 2010.
18. Wu, J., Sun JS., Song ML. & Liang L. A ranking method for DMUs with interval data based on DEA cross-eciency evaluation and TOPSIS, Journal of Systems Science and Systems Engineering, 22 (2), 191-201, (2013).
19. Yang, Z., Wang, X. & Sun, D. Using the bootstrap method to detect inuential DMUs in data envelopment analysis, Annals of Operations Research 173, 89-103, 2010.