Advances and Applications in Statistics
Volume 50, Issue 2, Pages 137 - 171
(February 2017) http://dx.doi.org/10.17654/AS050020137 |
|
RADIAL SENSITIVITY ANALYSIS OF LEFT, RIGHT, AND OVERALL RETURNS TO SCALE USING THE VIRTUAL ANTI-IDEAL UNIT
Mostafa Omidi, Mohsen Rostamy-Malkhalifeh, Ali Payan and Farhad Hosseinzadeh Lotfi
|
Abstract: Sensitivity analysis of the classification of returns to scale (RTS) and determination of the region wherein RTS classification of a decision making unit (DMU) remains stable provide managers with more accurate predictions regarding the changes to be exhibited by the targeted DMU. This study seeks to employ data envelopment analysis (DEA) to identify, locally, the regions having a stable RTS, and pursues this objective via a new approach to DMU performance analysis. Based on this new approach and using a different definition for left-RTS and right-RTS of frontier units, this study employs new DEA models to determine stability interval and stability radius of RTS classifications as well as the classifications of left-RTS, right-RTS and then overall-RTS of frontier units. The methods developed in this study can be easily applied to real world datasets by the use of existing DEA codes. |
Keywords and phrases: data envelopment analysis (DEA), left returns to scale (L-RTS), right returns to scale (R-RTS), stability, sensitivity, anti-ideal unit. |
|
Number of Downloads: 434 | Number of Views: 1532 |
|