【正文】
h in hypothesis,Alternative hypothesis(H1),4, 標(biāo)準(zhǔn)差的假設(shè)檢驗(yàn),1, 檢驗(yàn)兩組數(shù)據(jù)的標(biāo)準(zhǔn)差是否相等 2, 原假設(shè)和對(duì)立假設(shè): H0 σ1=σ2 ??H1 σ1σ2 3, 標(biāo)準(zhǔn):P 0.05時(shí), 兩組數(shù)據(jù)的標(biāo)準(zhǔn)差相等 4, Path in MINITAB: Stat?Basic statistics?Variances,假設(shè)檢驗(yàn),5,正態(tài)分布的假設(shè)檢驗(yàn),定義:檢驗(yàn)一組數(shù)據(jù)是否服從正態(tài)分布 假設(shè):H0: 正態(tài)分布??H1: 非正態(tài)分布 標(biāo)準(zhǔn):P 0.05時(shí), 數(shù)據(jù)為非正態(tài)分布 正態(tài)概率圖 計(jì)算平均值,標(biāo)準(zhǔn)差 將數(shù)據(jù)從小到大排序,計(jì)算各數(shù)據(jù)對(duì)應(yīng)的累積分布概率 描點(diǎn)(注意縱軸的刻度) Path in MINITAB: Stat?Basic statistics?Normality test,假設(shè)檢驗(yàn),5, 正態(tài)分布的假設(shè)檢驗(yàn),正態(tài)分布下的直方圖和正態(tài)概率圖,假設(shè)檢驗(yàn),5, 正態(tài)分布的假設(shè)檢驗(yàn),非正態(tài)分布下的直方圖和正態(tài)概率圖,假設(shè)檢驗(yàn),5, 正態(tài)分布的假設(shè)檢驗(yàn),在數(shù)據(jù)不服從正態(tài)分布時(shí),采用BoxCox變換改變數(shù)據(jù)的分布形狀 BoxCox變換 Y?Yλ (Path: Stat?control charts?BoxCox transformation…),假設(shè)檢驗(yàn),6,合格率的假設(shè)檢驗(yàn),類型 一批產(chǎn)品合格率是否小于P% 二批產(chǎn)品合格率是否相等 例 1,從生產(chǎn)產(chǎn)品中抽出2000進(jìn)行檢查,52不合格,合格率是否小于98%? 2,從一條生產(chǎn)線抽出1500產(chǎn)品檢查,17不合格;從另一條生產(chǎn)線抽出1300產(chǎn)品檢查,25不合格;它們的合格率是否一樣?,假設(shè)檢驗(yàn),6,合格率的假設(shè)檢驗(yàn),Proportion test in MINITAB,Proportion test for one group,Proportion comparison between two groups,假設(shè)檢驗(yàn),6,合格率的假設(shè)檢驗(yàn),例 1 (Proportion test for one group),輸入檢查結(jié)果,輸入檢驗(yàn)對(duì)比合格率,選擇假設(shè)類型,假設(shè)檢驗(yàn),6,合格率的假設(shè)檢驗(yàn),Test and CI for One Proportion Test of p = 0.98 vs p 0.98 Exact Sample X N Sample p 95.0% Upper Bound PValue 1 1948 2000 0.974000 0.979580 0.037,例 1 (Proportion test for one group),P0.05, 判斷結(jié)果合格率小于0.98,假設(shè)檢驗(yàn),6,合格率的假設(shè)檢驗(yàn),例 2 (Proportion comparison between two groups),假設(shè)檢驗(yàn),6,合格率的假設(shè)檢驗(yàn),例 2 (Proportion comparison between two groups),Test and CI for Two Proportions Sample X N Sample p 1 1483 1500 0.988667 2 1275 1300 0.980769 Estimate for p(1) p(2): 0.00789744 95% CI for p(1) p(2): (0.00129108, 0.0170860) Test for p(1) p(2) = 0 (vs not = 0): Z = 1.68 PValue = 0.092,P0.05, 判斷結(jié)果合格率相等。,假設(shè)檢驗(yàn),7,離散性數(shù)據(jù)相關(guān)性檢驗(yàn),例 缺陷嚴(yán)重度(數(shù)量)與加工速度關(guān)系 MINITAB: stat?tables?chisquare test… P0.05時(shí),兩個(gè)變量間有強(qiáng)的相關(guān)性。,假設(shè)檢驗(yàn),7,離散性數(shù)據(jù)相關(guān)性檢驗(yàn),ChiSquare Test: C1, C2, C3 Expected counts are printed below observed counts C1 C2 C3 Total 1 30 14 8 52 27.58 16.66 7.76 2 40 29 12 81 42