Performance Evaluation Considering Decision Preference

Extending conventional two-stage data envelopment analysis (DEA) models, this research presents novel multi-criteria decision-making approaches, which use DEA methods to measure the performance of two-stage production processes under different preference contexts, such as resource reallocation and group decisions. We consider two distinctly different decision preferences in our research. The first one considers preferences derived from decision makers, including their personal likes and dislikes, subjective weight settings, and so on; the second one, on the other hand, stems from the broad sense that decision preferences could arise from beyond decision makers’ discretions, such as from resource, cost, or environmental constraints, and more. This research closely examines the two-stage chain-structures of typical manufacturing procedures, and particularly focusing on how different managerial preferences affect decision strategies, decision sequences, and, ultimately, decision results. Our findings offer valuable insights into DEA-based multi-criteria decision making, and help businesses make wise managerial remedies based on different decision preferences settings.This six-chapter dissertation is organized as follows. Chapter 1 discusses basic data envelopment analysis theory and methods, and further introduces typical two-stage DEA models and their solutions; in Chapter 2, we study how decision preferences affect decision-making strategies under a single-stage DEA framework; in Chapter 3, we closely examine the chain structure of two-stage manufacturing procedures and their production possibility sets (PPS), based on which three PPS definitions are provided; Chapter 4 describes a novel CSDEA approach that employs frontier-shift methodology to measure ESC ecological efficiency when carbon permits are tradable, and that uses carbon trading quantitative information to determine to whom the trade should be made and how much to trade; and Chapter 5 proposes a novel multi-criteria group decision-making approach that uses CSDEA to appraise two-stage production processes under endogenous preferences. We apply the FSDEA model to the energy supply chain (ESC) so as to investigate how such decision preferences affect our decision methods, which helps validate our proposed model and develop managerial insights.The main contributions of this research are fourfold:1) We propose novel DEA models in which the decision making units (DMUs) compete, which means the total outputs of all DMUs are constrained. We first discuss the model under the circumstance of a fixed total amount of outputs, based on which we further establish models that could handle more general situations with ranges of different input and output variables. We find that the DMU under evaluation could optimize its performance either by increasing its outputs, or decreasing its inputs, and, more importantly, suggested adjustments would be no more, or worse, than the solutions suggested by conventional CCR models. Further, the proposed approach uses trading quantitative information to determine to whom the trade should be made and how much to trade, and reveals why certain DMUs are more likely involved in trading than others. These findings provide decision makers with managerial insights for making wise decisions in a competitive decision environment.2) We closely examine the two-stage structure and property of manufacture chains. We propose three PPS definitions for such two-stage procedures, and prove their equivalence. We further establish a novel PPS-based two-stage DEA model based on our findings. The technique proposed in the research efficiently determines the evaluation benchmarks for improving performance in series production systems, and provides decision makers with targeted managerial remedies.3) We propose a novel two-stage data envelopment analysis model (FSDEA) that employs frontier-shift methodology to measure energy supply chain (ESC) ecological efficiency when carbon permits are tradable. The proposed approach uses carbon trading quantitative information to determine to whom the trade should be made and how much to trade. It also reveals why certain decision making units (DMUs) are more likely to trade than others. Finally, the proposed model re-ranks all DMUs according to their ecological performance efficiency, as derived from the frontier-shift DEA model. FSDEA compares each DMU’s potential percentage improvement with minimal possible efficiency efforts (carbon reduction), which differs from extant DEA models that rank solely by efficiency scores. Our findings offer valuable insights into ESC efficiency and help businesses make better ecological decisions.4) This research presents a novel multi-criteria group decision-making approach that uses CSDEA to appraise two-stage production processes under two distinctly different decision strategies:efficiency- and fairness-based group decision preferences. Under CSDEA, we re-design the decision sequences, and examine how group-level managerial preferences impact multi-criteria group decisions outcomes, and, in turn, decision strategies. These findings offer valuable insights into DEA-based multi-criteria group decision-making, and help businesses make better collective decisions based on different group decision strategies priorities.