msa测量系统分析教材好文档.ppt

K Rater B = 数字 比例 计算分级员B的Kappa ? 利用Minitab的 二元分类 系数型 Gage RR的 Study Ex) 为了分析以Go/No-Go Gage判定合格/不合格的系数型测试系统 选了15 EA 样品 (合格品 8个,不合格品7个)之后由3名测试人员反复测了3次后得到了如下的Data. Attribute Gage RR Study Attribute Gage RR Study for Measurem Within Appraiser Assessment Agreement Appraiser # Inspected # Matched Percent (%) 95.0% CI A 15 14 93.3 ( 68.1, 99.8) B 15 13 86.7 ( 59.5, 98.3) C 15 13 86.7 ( 59.5, 98.3) # Matched: Appraiser agrees with him/herself across trials. ? 测量一致性 评价 -评价每个测试人员反复测试出的值之间的测量一致性 * 各测试人员的反复测量的结果的一致性的比率 反复测试每个样品时一致的样品数 总样品数 × 100 反复性 Each Appraiser vs Standard Assessment Agreement Appraiser # Inspected # Matched Percent (%) 95.0% CI A 15 14 93.3 ( 68.1, 99.8) B 15 12 80.0 ( 51.9, 95.7) C 15 12 80.0 ( 51.9, 95.7) # Matched: Appraisers assessment across trials agrees with standard. Assessment Disagreement Appraiser # NG/G Percent (%) # G/NG Percent (%) # Mixed Percent (%) A 0 0.0 0 0.0 1 6.7 B 1 6.7 0 0.0 2 13.3 C 0 0.0 1 6.7 2 13.3 # NG/G: Assessments across trials = NG / standard = G. # G/NG: Assessments across trials = G / standard = NG. # Mixed: Assessments across trials are not identical. ? 测量准确性(Accuracy)的 评价 -比较每个测试人员的测试值 vs 标准(参考值)时评价是否一致 * 每个测试人员反复测试结果的准确性 比率 标准值一致的次数 总样品数 ×100 有效性 误發警報率p(FA) 漏發警報率p(Miss) 点 :对比率的点推定值 线 : 信赖区间的宽度 * Graph 解释 ? 正确度 分析 测量已知参考值的15个部件时正确地判断了14次时,这时求正确判定的比率p的 点推定值 95信赖区间. ◎ Hint : STATBasic Statistics1Propotion Test and CI for One Proportion Test of p = 0.5 vs p not = 0.5 Exact Sample X N Sample p 95.0% CI P-Value 1 14 15 0.933333 (0.680515, 0.998314)