Finding Returns To Scale On Interval Data By Sensitivity Analysis In DMU

We recommend a method to find a return to a scale when outputs and inputs are assumed interval. Decision making units can be divided to three classes as follows: (i) efficie in any cases , (ii) efficient in best sense but for some data levels they lose their efficiency , (iii) always inefficient . We introduce models that show us how much we can move in direction increasing (decreasing) output (input) until fix type of return to scale (RTS). In the recent papers the issue of sensitivity analysis of data envelopment analysis (DEA) results has been extensively studied. Let where it is assumed lower and upper bounds of the intervals given are known and positive. The following model provides a lower bound for efficiency of .