The Bias in the DEA Two-stage Method (in Thai)
This paper studies the bias of using DEA two-stage method for measuring managerial efficiency. In using the method, first, the managerial efficiency is measured by a DEA method. Then, a tobit model is employed to examine the influence of an exogenous variable. Data used in this study were obtained with a Monte Carlo technique. The analysis determines the variance and the correlation of inputs and exogenous variable at different levels. The results show that the variance and the degree of correlation between inputs and exogenous variable significantly influenced the managerial efficiency score. This caused the efficiency score to be either too high or too low compared to the actual value. This also distorted the influence of exogenous variable, which thus caused a bias and gave an imprecise managerial efficiency score. The analyst thus should test the variance and correlation between inputs and exogenous variable before applying the DEA two-stage method. If problems are detected, this can be corrected by ridding the influence of exogenous variable from the output or reducing the variance of the data by using natural logarithm. These techniques would improve the precision of estimating managerial efficiency score.
Keywords: DEA two-stage method, bias evaluation, managerial efficiency
View Abstract 3809 Hit(s)
Download Fulltext 3739 Hit(s)