| Abstract: | 環境影響評估為保障環境與資源得以永續使 用之最基本手段之一,其評估過程可分為分析與綜 合兩部分,綜合評估之目的在提供一整體綜合評估 結果給決策者作參考,以決定開發計畫是否執行。 目前綜合評估方法中以加權比重計量法為研究之 主流,而權重更是綜合評估之重要步驟之一,然其權 重過程常牽涉心理計量問題,因此權重方法的選擇 與修正將是綜合評估成敗的重要關鍵。 Delphi法可以說是目前最適用、合宜的權重方 法之一,這是因其能使團體意見趨於一致,並可避免 面對面衝突及權威人士操縱的缺點。但其仍有可 能無法達到結果收斂的隱憂,亦即無法使團體達到 共識。因此本研究針對Delphi法無法收斂之問題加 以探討,應用統計分析方法中的因子分析方法來修 正Delphi法的運用步驟,以達加快收斂的目的。藉著 因子分析的特點---找出共通因子及與環境項目的 相關性,即那些環境因子在權重過程中是可以合併 考量的,將此分析的結果提供給參與者作參考,使參 與者可以有客觀的數據考量,以修正觀念達到共識 ,加快收斂速度。 本研究以馬鞍水力發電計畫為驗證案例,以馬 鞍計畫所用之Delphi法與本研究修正之Delphi法同時 進行權重,結果二者之權重結果差異不大、但本法 有提供客觀的統計資料,及提供是否繼續回饋的判 斷值,較原用之Delphi法為優。研究中除了證明修正 之Delphi法有加快收斂的優點之外,尚可將之應用於 評估範疇之界定及環境指標之製作。
Environmental Impact Assessment is well recognized as an effective planning tool in environmental protection and resource management. It includes two stages in the process, the analysis stage and the synthesis stage. The purpose of the synthesis process is to provide an evaluation for decision-makers. In the synthesis process, the weighting process is one of the major steps. But the weighting process is concerned with psychometric issues, therefore, the selection and modification of weighting method plays an crucial role in the evaluation process. Delphi method is one of the available weighting methods. It can avoid the disadvantages of face-to-face conflict and nonadjective decision. In addition, Delphi method can concur with the various opinions. Unfortunately, divers data are usually hard to converge. For the purpose of rapid convergency, the purpose of this study is to understand the divergency of Delphi method, and to implement the factor analysis of statistics to modify the Delphi method. Factor analysis can help us to understand the connectivity between common factors and environmental phenomenon. Thus, it can provide an useful reference for participators to modify their opinions towards accordant, and to reach the purpose of rapid convergence. In this study, the case of Ma-Ann hydropower plan was used. We compared the Delphi and the modified Delphi method used by Ma-Ann hydropower plan and this study, respectively. The weighting results of the two methods make no difference. But the modified Delphi method is better than the conventional Delphi method, because it can provide not only the adjective statistical data, but also the judgement if the feedback is continued. In addition to the advantages, the modified Delphi method can be applied on scoping and formation of environmental indices. |
