An approach of linear goal programming with priority for improving weigth dispersion in the data envelopment analysis

Data Envelopment Analysis (DEA) has a widespread usage method for measuring and bencmarking relative efficiency of peer decision making units (DMUs) with multiple inputs and outputs. Beside of its widespread usage, DEA has some drawbacks, the most known of which is the unrealistic weight dispersion. Although this case is unreasonable and undesirable, unrealistic weight dispersion occurs, when some DMUs are rated as efficient because of input and output weights have the extreme or zero values. DMUs which have extreme values have potential to be a false positive. The solution of proposed Multiple Criteria Data Envelopment Analysis by Linear Goal Programming with priority, produces more homogeneous weight dispersion for inputs-outputs without a priori information and without any additional constraints on weights.