不可儲存商品之生產績效衡量--整合式資料包絡分析模式

Abstract

在評估不可儲存商品（如運輸服務）之生產績效時，由於生產效率與生產效果之績效值並不等同，故應同時衡量兩部分，方不至於偏頗。當不可儲存商品一旦生產且有一部分產出未能同時被消費時，則其生產效果(指技術效率與服務效果的綜合效果)的績效值將小於生產效率的績效值。有鑑於此，本研究試圖建構一整合式資料包絡分析 (Integrated Data Envelopment Analysis; IDEA) 模式，以同時求解技術效率及服務效果的績效值。本IDEA 模式可同時決定投入變數、生產變數及消費變數的參數值，亦可證明具有「合理性」及「唯一性」。經個案分析結果顯示，本IDEA模式的鑑別力較傳統DEA模式高。另，本文進一步建構一般化IDEA模式，可看出權重變動對各個DMU之影響，俾提出更多改善效率之方法。

Efficiency and effectiveness for non-storable commodities such as transport services represent two distinct measurements. When such commodities are produced and a portion of which are not consumed instantaneously, the technical effectiveness (a combined effect of technical efficiency and service effectiveness) would be likely less than the technical efficiency. Based on this, this thesis attempts to develop an integrated data envelopment analysis (IDEA) model that can jointly determine the overall efficiency from the aspects of technical-efficiency, service-effectiveness, and technical-effectiveness. The core logic for the proposed IDEA model is to simultaneously determine the virtual multipliers associated with the variables of factor production and consumption. The underlying properties of reasonability and uniqueness of the proposed IDEA model are proven. The applicability of the proposed model is also demonstrated with a case study. It shows that our proposed IDEA model has higher discrimination power than the conventional separated DEA models.