應用資料包絡分析法（DEA）於六標準差專案之績效評估

Performance Evaluation of Six Sigma Projects – An Data Envelopment Analysis Approach

目前全球企業在六標準差改善專案整體流程導入及工具手法之應用，皆趨於成熟階段，但在各企業成功導入六標準差專案的同時，卻缺乏一架構完整之績效評估模式。本研究目的在運用資料包絡分析法 （Data Envelopment Analysis,DEA） 分析評估三十則國內外近三年內所推導六標準差專案，含流程改善（IFSS） 與設計改善 （DFSS） 執行績效。研究中利用DEA 模式，首先由CCR模式求得整體相對效率，再以BCC 模式求得技術效率值與規模效率值，並透過差額變數分析求取在總生產效率下各專案最適投入變數，最後運用敏感度分析，探討當專案投入項與產出項減少時，對相對效率值之影響。研究結果提供受評業者進行六標準差專案規劃、決策時之參考依據。

The methodologies and implementation of six sigma management hasbecome mature worldwide. However, there is still lack of subjective andstandard methods to evaluate the performance of six sigma project. Based on the Data Envelopment Analysis (DEA), this research is to develop aprocedure to explore the performance of six sigma projects includingImprovement for Six Sigma (IFSS) and Design for Six Sigma (DFSS).Atotal of thirty projects are considered in this research. The CCR model isemployed to obtain the overall relative effectiveness. BCC model is thenused to further evaluate the technical effectiveness and scale effectiveness.Through the slack variable analysis, the optimum input can be determined.Finally, the sensitivity analysis helps explore the influence from theelimination of individual input parameter.