【正文】 ? 執(zhí)行的時間 Timing to implement ? 解決方案的控制度 Control over the solution ? 解決方案的成本節(jié)約 Cost savings for that solution ? 客戶風險 Risk to the customer ? 資源需求 Resources required ? 業(yè)務影響 Business impact ? 解決方案的難度 Difficulty of the solution 客戶 (管理層 )是制定標準的一個好來源 The customer (management) is a good source for the criteria. 解決方案矩陣 Solution Matrix DA 4 Basic Tools Basic Tools Six Sigma Training 48 加權解決方案抉擇矩陣的步驟 Steps to create a Weighted Solution Selection Matrix 2) 確定范圍內(nèi)的每個風險標準的相關權重 . 通常采用數(shù)值 13 , 15, 和 110. 最高 的數(shù)值用在最重要的標準上 . Define the relative weighting of each risk criteria in your area. Commonly used scales are 1 to 3, 1 to 5, and 1 to 10. Use the higher end of the scale for the most important criteria. ? 如果可能 , 采用 絕對 標準 (小于 , 大于 , 等于 ) If possible, try to use ABSOLUTE Criteria (less than, greater than, shall have, etc.) ? 依據(jù)一個標準 , 絕對 方法對選擇解決方案非常有效 . ABSOLUTES are good for ruling in and out potential solutions based upon just one criteria 3) 列出需要研究的所有解決方案 List all solutions that need to be studied. 解決方案矩陣 Solution Matrix DA 4 Basic Tools Basic Tools Six Sigma Training 49 解決方案矩陣 Solution Matrix 加權解決方案抉擇矩陣的步驟 Steps to create a Weighted Solution Selection Matrix 4)針對全部標準給每個解決方案評分 . 通常通常采用數(shù)值 13 , 15, 和 110. 最高的數(shù)值用在最想采用的行為上 .. Grade each solution with respect to all criteria. Commonly used scales are 1 to 3, 1 to 5, and 1 to 10. Use the higher end of the scale for the most desirable performance. 5) 將每個標準的權重乘以行為分數(shù)得到標準分 . Multiply each criteria weighting by the performance score to calculate the criteria score.
?；貧w分析按照涉及的變量的多少，可分為一元回歸分析和多元回歸分析；按照自變量和因變量之間的關系類型，可分為線性回歸分析和非線性回歸分析。 Dot Plots) ? 用圖形反映一個流程中的兩個變量之間可能存在或發(fā)生的關系 . A graphical representation of possible relationships between two variables in a process. ? 分布在圖中的圓點是數(shù)據(jù)點 . Dots representing data points are scattered on the diagram ? 每一群叢生在一起的圓點形成的一條線顯示了兩個因素的強相關性 The extent to which the dots cluster together to form a line across the diagram show the strength of correlation between the two factors ? 原因和結(jié)果不一定是直線關系 – 可能有第三個無法圖示的變量作用于這兩個變量 Note that a straight line does not necessarily indicate cause and effect – both variables may be driven by an unplotted third variable DA 4 Basic Tools Basic Tools Six Sigma Training 34 散點圖 Scatter Diagram 散點圖說明 Interpreting Scatter Diagrams ? 查看圖的全貌 Look for a general pattern of plotted points ? 數(shù)據(jù)聯(lián)系的緊密程度顯示有相關性 Tightness of grouping indicates relative correlation Positive Correlation 正相關 Negative Correlation 負相關 No Correlation 沒有相關性 DA 4 Basic Tools Basic Tools Six Sigma Training 35 如何繪制散點圖 How to construct a Scatter Diagram: ? 確定驗證的一對要素 Decide which paired factors to examine. ? 根據(jù)經(jīng)驗 , 最少的數(shù)據(jù)點是 30個 30 points is a good ruleofthumb for the minimum number of points to use. ? 根據(jù)你的情況做調(diào)整 Adjust to your particular situation. ? 找出每個變量的最大值和最小值 ,確定坐標軸的數(shù)值范圍 Find the highest and lowest values for each variable to set the scale of the axes. ? 畫圖 , 給圖加標題 , 標注坐標軸 Plot the data, titling the plot and labeling the axes. DA 4 Basic Tools Basic Tools Six Sigma Training 36 散點圖 – 練習 Scatter Diagram Exercise 以小組形式 , 根據(jù)下面給出的數(shù)據(jù)繪制 散點圖 As a team, create a scatter diagram on a flip chart based on the following dataset: A B19 2322 3022 2926 329 204 727 2729 3810 1424 2610 1117 3027 2914 1612 1920 262 91 1321 277 91)有相關性嗎 ? 1) Is there correlation? ? 如果有 , 是正相關還是負相關 ? If yes, is it positive or negative? 2)是否驗證了原因和結(jié)果 ? 2) Does this prove Cause and Effect? DA 4 Basic Tools Basic Tools Six Sigma Training 37 ? 相關性指明了兩個變量之間的關系 Correlation indicates a relationship between two variables ? 相關并不是指明兩個變量的因果關系 Correlation does not indicate cause and effect ? 相關性的強弱用 R來表示 The strength of a correlative relationship is identified by R Rules of thumb: or indicate a strong linear correlation between the two variables 相關性 Correlation DA 4 Basic Tools Basic Tools Six Sigma Training 38 R又被稱作相關系數(shù) R Is The Correlation Coefficient R = R = 0 R = + R = +1 1 R +1 R = 1 相關性 Correlation DA 4 Basic Tools Basic Tools Six Sigma Training 39 回歸 Regression 相關和回歸 Correlation And Regression ? 相關并不是指兩者的因果關系 Correlation does not indicate cause and effect ? 變量可以相互關聯(lián)地變化但不一定是因果關系 Variables can covary without one causing another ? 若要分析兩個變量之間是否存在因果關系，我們需要用回歸的方法進行分析 When analysis is looking for cause and effect, regression is used DA 4 Basic Tools Basic Tools Six Sigma Training 40 Regression ? Regression looks to the specifics of the cause and effect relationship between continuous variables ? Regression creates a predictive equation ? Regression assesses the strength of the relationship using R2 回歸分析是研究一個變量 Y與其它若干變量 X之間相關關系的一種數(shù)學工具，它是在一組試驗或觀測數(shù)據(jù)的基礎上的，尋找被隨機性掩蓋了的變量之間的依存關系。它是揭示過程輸出缺陷或問題與其潛在原因之間關系的圖表，也是表達和分析其因果關系的重要工具和文檔。 將結(jié)果記分 , 去掉最低分 , 重復上述步驟減少清單上的項目 . Option 2: Weighted voting – each team member distributes x points among the items on the list。 repeat the process with the reduced list util an agreed upon list is generated. DA 4 Basic Tools Basic Tools Six Sigma Training 14 多層次投票法 MultiVoting 多層次投票的選擇項 – 排序 或加權 Options for MultiVoting Force Rank or Weight... ? 選擇 1: 排序投票 – 團隊成員依照重要程度寫下 n個項目的排序 ,每個成員都要打分 , 最重要的項目分值是 n,第二重要的項目分值為 n1,依此類推 . Option 1: Rank Voting each team members writes their n items in order of importance。t worked topic topic DA 4 Basic Tools Basic Tools Six Sigma Training 12 頭腦風暴 Brainstorming 現(xiàn)在團隊有了他們小組由頭腦風暴發(fā)掘的建議清單 …如何找出團隊一致同意的最重要的建議呢 ?我們不可能調(diào)查所有的問題 … 我們怎樣處理這種情況呢 ? Now the team has a Brainstorming list of ideas in their natural groups... how can the team agree on what ideas