A statistical test for outlier identification in data envelopment analysis

Subject Areas : Data Envelopment Analysis

Morteza  Khodabin ¹ (Islamic Azad University, Karaj Branch, Department of Mathematics P.O.Box 31485-313, Karaj, Iran.)
Reza  Kazemi Matin ² (Islamic Azad University, Karaj Branch, Department of Mathematics P.O.Box 31485-313, Karaj, Iran.)

Received: 2010-08-30 Accepted : 2010-08-30 Published : 2010-09-01

Keywords: Data Envelopment Analysis (DEA), Efficiency, Outlier, Paired Sample Test,

Abstract :

In the use of peer group data to assess individual, typical or best practice performance, the effective detection of outliers is critical for achieving useful results. In these ‘‘deterministic’’ frontier models, statistical theory is now mostly available. This paper deals with the statistical pared sample method and its capability of detecting outliers in data envelopment analysis. In the presented method, each observation is deleted from the sample once and the resulting linear program is solved, leading to a distribution of efficiency estimates. Based on the achieved distribution, a pared test is designed to identify the potential outlier(s). We illustrate the method through a real data set. The method could be used in a first step, as an exploratory data analysis, before using any frontier estimation.

References:

[1] Charnes A., Cooper W.W., Rhodes E., Measuring the efficiency of decision making units. European Journal of Operational Research, 2, 29-444, 1978.