theyieldofGrindBeadproducts-GB六西格码项目.ppt

Continuous Improvement 31 KPIV 2: Change the KPIV2 can impact the Chip scrap rate. ? KPIV 2 Detail Firing process — Top-temperature Before: 1220 ℃ for 24B0110-010 Change method: 1190 ℃ for 24B0110-010 ? Hypothesis Test ? Two Samples test will be used to test this KPIV. We need to sure temperature change will not affect the final properties. ? Hypothesis Statement ? Ho : After change KPIV1, the properties are as same as before. ? Ha : After change KPIV1, the properties scrap rate are not same as before. Continuous Improvement 32 ? Power and Sample Size for R and Φ . ? Testing mean 1 = mean 2 (versus not =) ? Calculating power for mean 1 = mean 2 + difference ? Alpha = 0.05 Assumed standard deviation = 0.3857 ? Sample Target ? Difference Size Power Actual Power ? 0.05 1548 0.95 0.950062 ? The sample size is for each group. We need 1548 samples data for this contrast at least. KPIV 2: Change the KPIV2 can impact the Chip scrap rate. We changed the KPIV2 on Oct-2010. Continuous Improvement 33 Hypothesis Test (Chip ) ? P-Value=0,378 P(1) =P(2) ? P-Value=0, 883 P(1) =P(2) Comments : P-Value 0.05 , the result shows after reduced the top temperature, Properties ( R and Φ ) is still same with before. The properties didnt change so much. Two-sample T for 1190-2 vs 1220 N Mean StDev SE Mean 1190-2 25 13.516 0.395 0.079 1220 25 13.449 0.386 0.077 Difference = mu (1190-2) - mu (1220) Estimate for difference: 0.066400 95% CI for difference: (-0.155591, 0.288391) T-Test of difference = 0.05 (vs not =): T-Value = 0.15 P-Value = 0.883 DF =48 Both use Pooled StDev = 0.3904 Individual Value Plot of 1190-2, 1220 Boxplot of 1190-2, 1220 KPIV 2: Change the KPIV1 can impact the Chip scrap rate. Two-sample T for 1190-1 vs 1220 N Mean StDev SE Mean 1190-1 25 13.423 0.189 0.038 1220 25 13.449 0.386 0.077 Difference = mu (1190-1) - mu (1220) Estimate for difference: -0.026400 95% CI for difference: