Abstract
Aim/purpose ‒ The traditional data envelopment analysis (DEA) is popularly used to evaluate the relative efficiency among public or private firms by maximising each firm’s efficiency: the decision maker only considers one decision-making unit (DMU) at one time; thus, if there are n firms for computing efficiency scores, the resolution of n similar problems is necessary. Therefore, the multi-objective linear programming (MOLP) problem is used to simplify the complexity.
Design/methodology/approach ‒ According to the similarity between the DEA and the multiple attribute decision making (MADM), a game of MADM is proposed to solve the DEA problem. Related definitions and proofs are provided to clarify this particular approach.
Findings ‒ The multi-objective DEA is validated to be a unique MADM problem in this study: the MADM game for DEA is eventually identical to the weighting multi-objective DEA. This MADM game for DEA is used to rank ten LCD companies in Taiwan for their research and development (R&D) efficiencies to show its practical application.
Research implications/limitations ‒ The main advantage of using an MADM game on the weighting multi-objective DEA is that the decision maker does not need to worry how to set these weights among DMUs/objectives, this MADM game will decide the weights among DMUs by the game theory. However, various DEA models are eventually evaluation tools. No one can guarantee us with 100% confidence that their evaluated results of DEA could be the absolute standard. Readers should analyse the results with care.
Originality/value/contribution ‒ A unique link between the multi-objective CCR DEA and the MADM game for DEA is established and validated in this study. Previous scholars seldom explored and developed this breathtaking view before.