The model has input and output variables that were mean normalized by dividing the respective observations of the variable by its mean.
There after, I ran an input-oriented DEA with variable returns-to-scale. The resulting VRS_TE output in STATA had value range between 0 and 1, so yay!
But then I read someplace that the STATA output of DEA scores are the reciprocal and ranges from 1 to infinity, thus requiring some transformation in order to get to the popular efficiency scores (between 0 and 1), which is done by dividing 1 by the STATA output. However, the DEA results STATA yielded already ranged 0 to 1.
Do I no longer need to transform the STATA results? If so, may I ask for a simple explanation on how this came about?
My apologies for my shallow understanding of DEA. Any help on this will be very much appreciated.
There after, I ran an input-oriented DEA with variable returns-to-scale. The resulting VRS_TE output in STATA had value range between 0 and 1, so yay!
But then I read someplace that the STATA output of DEA scores are the reciprocal and ranges from 1 to infinity, thus requiring some transformation in order to get to the popular efficiency scores (between 0 and 1), which is done by dividing 1 by the STATA output. However, the DEA results STATA yielded already ranged 0 to 1.
Do I no longer need to transform the STATA results? If so, may I ask for a simple explanation on how this came about?
My apologies for my shallow understanding of DEA. Any help on this will be very much appreciated.
