1、2022-10-1Minitab使用教學2022-10-1目目 錄錄0.平均數與標準差圖示平均數與標準差圖示1.平均數與標準差平均數與標準差2.常態檢定常態檢定3.直方圖直方圖4.箱型圖箱型圖6.Run Chart5.二項分配二項分配2022-10-1 (u)MEAN(群體平均)TIGHT DISTRIBUTION(緊密分配)(很小)BROAD DISTRIBUTION(很大)變異-3 -2 -1 TARGET +1 +2 +3 ：是測量離勢或變異(變異情形通常以平均值和標準差來表示品質特性值之位置及分散情形)其值愈小愈趨中心分配，品質變異越小越穩定，亦即是平均值愈接近目標值。反之，愈大分配離
2、中心值愈遠品質愈壞。NXXXnxNX2)(1)(2nXX+-:母群體平均數(趨中勢)(=):樣本平均數(品質特性平均值之估計值)(=):(離中趨勢)標準差(母群體)(=)S:樣本標準差(品質特性之估計值)(S=)常態分配機率平均數及標準差平均數及標準差2022-10-11.平均數&標準差A.選擇Stat Basic Statistics Display Descriptive Statistics.2022-10-11.平均數&標準差B.在Variable選入”C1”。檢查 By Variable選入C2C.選擇OK 2022-10-1平均數標準差2022-10-172.常態分配檢定A.選擇S
3、tat Basic statistics Normality Test.B.在Variable 選入”Height”。C.選擇OK HeightPercent807570656099.99995908070605040302010510.1Mean0.05068.72StDev3.659N92AD0.748P-ValueProbability Plot of HeightNormal P Value需大於0.052022-10-13.直方圖-1A.選擇Graph Historgram With Fit.B.在Graph Variable選入”Height”。C.選擇OK HeightFrequ
4、ency7572696663601086420Mean68.72StDev3.659N92Histogram of HeightNormal 可看到平均數、標準差2022-10-13.直方圖-2-1A.選擇Stat Quality Tools Capability Analysis Normal.2022-10-13.直方圖-2-2B.如下圖輸入。輸入要計算的數據行輸入Group的大小值(建議值：5)2022-10-13.直方圖-2-3C.在Lower spec/Upper spec Key 入規格值。輸入上下限規格值2022-10-13.直方圖-2-4D.進入 Option 在”Key 入中
5、心規格 按OK OK規格中心值或目標值規格中心值或目標值2022-10-1360350340330320310300LSLTargetUSLProcess DataSample?N39StDev(Within)5.12353StDev(Overall)4.97411LSL300Target330USL360Sample Mean309.282Potential(Within)CapabilityCCpk1.95Overall CapabilityPp2.01PPL0.62PPU3.40PpkCp0.62Cpm0.461.95CPL0.60CPU3.30Cpk0.60Observed Perfo
6、rmancePPM?USL0.00PPM?Total0.00Exp.Within PerformancePPM?USL0.00PPM?Total35020.06Exp.Overall PerformancePPM?USL0.00PPM?Total31015.49WithinOverallProcess Capability of C13.直方圖-2-5估計驗收組內變異整個系統有持續性系統改善組間變異2022-10-14.箱型圖A.選擇Graph Boxplot2022-10-14.箱型圖B.依需求選擇所要的箱型圖模式2022-10-14.箱型圖C.在Graph Variables輸入要分析的數
7、據列 OK2022-10-14.箱型圖D.圖形如下，有一異常點C1340330320310300Boxplot of C1異常點2022-10-15.二項式分配異常點A.選擇Calc Probability Distribution Binomial.2022-10-15.二項式分配B.如為機率分配選擇 Probability 輸入樣本數輸入機率輸入要求數2022-10-15.二項式分配C.答案如下：2022-10-15.二項式分配D.如為累積分配選擇 Cumulative probability2022-10-15.二項式分配E.結果如下：2022-10-16.Run ChartA.選擇Sa
8、ta Quality tool Run chart2022-10-16.Run ChartB.如下輸入2022-10-16.Run ChartC.結果如下圖：SampleC135302520151051340330320310300Number of runs about median:0.2866820Expected number of runs:18.37143Longest run about median:4Approx P-Value for Clustering:0.71332Approx P-Value for Mixtures:Number of runs up or down:0.7948621Expected number of runs:23.00000Longest run up or down:3Approx P-Value for Trends:0.20514Approx P-Value for Oscillation:Run Chart of C12022-10-17.Gage R&R2022-10-1
