應用Kappa分析法評估目視人員檢驗之一致性

Abstract

隨著國際市場的競爭日趨激烈，客戶對品質的要求也更趨嚴格，必須靠精準的量測系統，才能保證產品於製程中及出貨品質的可靠性，對於需要高度仰賴人工目視檢驗產品的產業而言，如何判定目視檢驗人員對判定產品好壞之能力是項非常重要的問題，因此本研究以計數型GO/NOGO量測系統分析(Measurement System Analysis)MSA為主題，應用Kappa分析法針對一個半導體測試廠，將目視人員檢驗結果透過Kappa分析法之交叉表計算出Kappa值，以探討目視人員彼此間對相同產品之品質特性進行好壞判定時，其判定結果的一致性，以及探討各目視人員判定結果對標準值之一致性與有效性，進而了解誤警率(將好的判定成不好的機率)及失誤率(將不好的判定成好的機率)，最後依據判定準則確認誤判風險是否可被接受，同時從分析中了解訓練目視人員時所需加強之方向，並給予不合格人員適當的教育訓練。 本研究最後以新竹科學園區某半導體測試廠為例，來說明本研究方法之有效性，研究成果可提供業者作為判定目視檢查人員對產品好壞判定之一致性與正確性能力之風險管理的參考，以期提高目視人員檢驗水準、減少誤判、有效攔阻不合格品流出及降低內外部品質成本損失，使企業在管理檢驗人員變異及訓練上有一套具體的做法。

Due to the serious competition of the international market, requirements of product /process quality from customers are increasing rapidly. Consequently, on the precise measuring system to ensure the product/process quality depends the Semiconductor manufactories which particularly needs precise visual inspection to ensure their product quality, therefore evaluating the agreement among inspectors, This study develops a procedure of assessing the attribute agreement of inspectors using Kappa becomes important analysis method. Finally, a case of IC testing house in Hsin-chu, Taiwan is utilized to demonstrate the effectiveness of the proposed method. The result of this study can be utilized by Semiconductor manufactories to assess their inspectors’ skills. it can also be employed as risk management. In addition, the results of this study can be utilized to reduce the quality lost of a process and to develop a useful method of improving inspector’s skills.