A modi ed test for detecting in uential decision-making units in data envelopment analysis

Abstract

In data analyses based on a deterministic or stochastic approach, usingpre-study is very important to identify observations that are not suitableto data in general. Among such observations, those that have ahigh tendency to change results negatively are called inuential observations.In this paper, we propose a new method to identify inuentialobservations in Data Envelopment Analysis (DEA). Our method is amodied version of the one proposed by Pastor et al. [12]. Both methodsare compared by using two well-known data sets and the outcomesare discussed. A comparative analysis indicates that our method isan eective alternative to the Pastor et al. [12] method to identifyinuential observations in DEA.