資料包絡分析法與比例分析法運用於科技專案效率評估之研究

Abstract

本研究主要是以資料包絡分析法（Data Envelopment Analysis）與比 例分析法(Ratio Analysis)來評估科技專案的執行效率，且以經濟部所屬 之科技專案作為實證的資料。由於科技專案的性質是屬於多投入與多產出 ，且各投入產出項之間並不容易確認、定義與衡量；從資料包絡分析法與 比例分析法的文獻探討中，發現此二方法可以適用於科技專案的效率評估 模式，藉以建立一個完整的評估科技專案效率的流程與模式。 本研究 所進行之評估分析，係就同一年度科技專案的投入產出資料進行檢驗與篩 選，選擇出符合執行效率評估所需之資料，接續進行資料包絡分析法與比 例分析法之評估活動：在資料包絡分析法之評估活動中，首先是利用CCR 模式:導出各受評計畫總效率值即評估出各計畫是屬於相對無效率群及相 對有效率群，並對相對無效率群進行差額分析，以進行改善。另進一步利 用BCC模式:藉以導出純技術效率及規模效率；使對相對無效率群進行規模 報酬之判定，以作進一步的分析與評估。而在比例分析法之評估活動中， 首先是建立並導出各評估指標及總結評估指標，接著利用總結評估指標以 評估各受評計畫之優劣順序，以評比各計畫之執行效率；此外並提供各計 畫對目標設定的參考資訊。在完成兩評估方式之評估活動後，就兩評估方 式之結果作比較；利用統計方法衡量其間之相關性與一致程度。並進而結 合兩評估方式，建立一整合的評估模式，並進行兩模式結合應用之探討。 藉由以上分析，在資料包絡分析法方面，評估出有效率的執行單位和無效 率的執行單位，並建議無效率單位改進的方向和大小；而利用比例分析法 ，將可評估出各計畫其於各評估指標及總結評估指標中，其執行效率之情 形，並可描述各投入產出間絕對的比例，提供計畫管理者判斷各投入產出 指標的重要性。而在資料包絡分析法與比例分析法評估之結果，頗為一致 。且經由兩模式的結合，可發現其兩方法可為互補，使計畫管理者利用於 各計畫之執行上資源分配之參考，即利用資料包絡分析法針對相對無效率 群提供改善之方式，而以比例分析法可提供相對有效率群於資訊於未來之 目標設定。即可提供專案執行者和管理當局管理和控制的參考。 This research is base on Data Envelopment Analysis (DEA) and RatioAnalysis as tools for assessing the executive efficiency of R&D Projects. The assessment of Ministry of Economic Affairs in Taiwan on the provision of technology projects is used as a vehicle for the two methods. Such projects typically use one or more resources to secure one or more output, the inputsand outputs being possibly incommensurate. With the literature review of theDEA and Ratio Analysis is suited for assessing the R&D project and is thereforebuild and an integrated framework to evaluate R&D projects is also produced. The process of evaluation in this research, first, adopting empirical datato make it possible to identify in the succeeding section input and outputvariable which can be used to assess productive efficiency in R&D projectswithin the same year.Second, using this two methods to assess. In DEA assessingactivity , indicates the relative total efficiency and will enable to find themost productive scale size and employ the slack variable analysis to discussthe operation status and level of improvements in the CCR model. To evaluateR&E performance to determent the technical efficiency and scale efficiency ,furthermore , in the BCC model. In another side, in Ratio Analysis method toassess the performance will construct PIs( Performance Indicators) Defined asthe ratios of each output to each input variable used within the DEA model .And construct the summary PIs to be intended to convey, in summary form, theoverall performance of the R&D projects. After the evaluated activity, thecomparison of two methods focuses on how well the two methods agree on theperformance of a project relative to that of other units, and on the estimatesof targets each method provides for improving the performance of projects.Finally, to build a combination of evaluating model to discuss and analysis. By used the analysis above, the two methods were contrasted in two aspect;the measure or measures of performance and the targets they provide. DEA is to evaluate the performance of the technology projects, and suggest the direction for inefficient units. Ratio Analysis is to evaluate the overall performance in DEA-efficient units. The combined DEA and Ratio Analysis information, reinforced by judgement, can prove useful for indicating the areas where a relatively efficient projects may strengthen its performance further, even if precise improved input and output levels cannot be estimated for a project already DEA efficient. Using these analyses as a reference to those project executives and administrators for controlling and management the performance of R&D projects.