Minitab的基础统计

附录1.利用Minitab的基础统计LGENTSIXSIGMATASKTEAM基础统计BeforeAfter58.560.060.354.961.758.169.062.164.058.562.659.956.754.4例题1:求基本的统计量•开发新产品的公司，为了知道它的效果，把7名主妇为对象:实验结果他们的体重变化如右图为了知道体重变化的程度？•Stat/basicstatistics/displaydescriptivestatistics附录1-1/41LGElectronics/LGENT6σTASKTEAM基础统计DescriptiveStatisticsVariableNMeanMedianTrMeanStDevSEMeanBefore761.8361.7061.834.011.52After758.2758.5058.272.791.05VariableMinimumMaximumQ1Q3Before56.7069.0058.5064.00After54.4062.1054.9060.00•TrimmedMean(整理平均)Minitabremovesthesmallest5%andthelargest5%ofthevalues(roundedthenearestinteger),andthenaveragestheremainingdata•SEMean(StandardErrorofMean)StDev/N例题1:求基本的统计量附录1-2/41LGElectronics/LGENT6σTASKTEAM基础统计•calc/calculatorBeforeAfter58.560.02.564160.354.9-8.955261.758.1-5.834769.062.1-10.000064.058.5-8.593862.659.9-4.313156.754.4-4.0564附录1:求基本统计量附录1-3/41LGElectronics/LGENT6σTASKTEAM基础统计•Random抽取工厂所生产的21个产品后，测定结果如下.求例题1中求的统计量外的多种的统计量？•calc/columnstatisticsorrowstatistics•stat/basicstatistics/storedescriptivestatistics例题2:求多种统计量附录1-4/41LGElectronics/LGENT6σTASKTEAM基础统计•执行这样的过程，计算的DataWindow以列别出现.例题2:求多种统计量附录1-5/41LGElectronics/LGENT6σTASKTEAM基础统计例题2:求多种统计量•统计量要在SessionWindow上显示?•Manip/displaydata选择所愿的统计量DataDisplayRowsizeMean1SEMean1StDev1Variance1Q1_1Median1Q3_1127.528.21900.4244671.945153.7836226.928.429.65227.6327.6430.3528.8622.9726.6831.8928.41026.91130.01231.21329.41428.01526.81628.81728.51826.31929.92026.92128.4RowIQR1Sum1Minimum1Maximum1Range1N112.75592.622.931.88.921DataDisplayRowMean1SEMean1StDev1Variance1Q1_1Median1Q3_1IQR1128.21900.4244671.945153.7836226.928.429.652.75RowSum1Minimum1Maximum1Range1SSQ11592.622.931.88.916798.3附录1-6/41LGElectronics/LGENT6σTASKTEAM基础统计例题3:概率分布的活用-2项分布(1)•感染到某种传染病能恢复的概率为0.4。15名被传染时，5名恢复的概率为?•Calc/probabilitydistributions/binominal试行成功概率成功次数ProbabilityDensityFunctionBinomialwithn=15andp=0.400000xP(X=x)5.000.1859附录1-7/41LGElectronics/LGENT6σTASKTEAM基础统计例题3:概率分布的活用-2项分布(2)•感染到某种传染病能恢复的概率为0.4.在15名当中至少10名能恢复的概率为？•Calc/probabilitydistributions/binominal执行次数成功概率成功次数DataDisplayK20.0338333*在Minitab上，为了求P(Xx)求P(X≤x)的值后计算1-P(X≤x)，求概率.附录1-8/41LGElectronics/LGENT6σTASKTEAM基础统计例题3:概率分布的活用-泊松分布•显示器的某种特性服从平均为10的泊松分布时，求P(X≤14)?•Calc/probabilitydistributions/PoissonCumulativeDistributionFunctionPoissonwithmu=10.0000xP(X=x)14.000.9165附录1-9/41LGElectronics/LGENT6σTASKTEAM基础统计例题3:概率分布的活用-正态分布•显示器的某种特性为正态分布时，求P(X≤-0.28)?•Calc/probabilitydistributions/normalCumulativeDistributionFunctionNormalwithmean=0andstandarddeviation=1.00000xP(X=x)-0.28000.3897附录1-10/41LGElectronics/LGENT6σTASKTEAM基础统计例题4:母平均的置信区间-知道母分散时•从分散为225的正态分布母总体取出25个样本，得到如下资料,样本的平均为64.32。求出母平均的95%的置信区间•Stat/basicstatistics/1SampleZ55697847495285733371708976555153527187387067817363ZConfidenceIntervalsTheassumedsigma=15.0VariableNMeanStDevSEMean95.0%CIC12564.3215.113.00(58.44,70.20)附录1-11/41LGElectronics/LGENT6σTASKTEAM基础统计例题4:母平均的置信区间-不知道母分散的情况•推定新显示器的寿命时间，抽取9台，寿命时间以小时为单位测定的结果如下假设寿命时间为正态分布时，求母平均寿命的90%的置信区间？•Stat/basicstatistics/1Samplet500005100054000520005400050000530005200052000TConfidenceIntervalsVariableNMeanStDevSEMean90.0%CIC29520001500500(51070,52930)附录1-12/41LGElectronics/LGENT6σTASKTEAM基础统计例题5:母平均差异的置信区间-两个母分散相同情况•对两个公司的HingeForce差异的分析，取出各20个样本的结果如下.求两个公司HingeForce差异的95%置信区间(假设两个母分散是一样的)•Stat/basicstatistics/2SampletPK2RPM12.072.172.132.172.172.152.072.162.112.122.062.132.122.132.142.142.12.112.112.172.162.112.082.122.122.192.092.122.182.152.112.142.192.122.182.152.132.122.122.12TwoSampleT-TestandConfidenceIntervalTwosampleTforPKvsRPMNMeanStDevSEMeanPK202.12200.03870.0087RPM202.13950.02330.005295%CIformuPK-muRPM:(-0.0380,0.0030)T-TestmuPK=muRPM(vsnot=):T=-1.73P=0.091DF=38BothusePooledStDev=0.0320假设母分散是一样的附录1-13/41LGElectronics/LGENT6σTASKTEAM基础统计例题5:母平均差异的置信区间-两个母分散不同的情况•为了知道对两个公司的HingeForce的差异，推出各20个标本的结果.对两个公司HingeForce的差异，求出95%的置信区间•Stat/basicstatistics/2SampletPK2RPM12.072.172.132.172.172.152.072.162.112.122.062.132.122.132.142.142.12.112.112.172.162.112.082.122.122.192.092.122.182.152.112.142.192.122.182.152.132.122.122.12TwoSampleT-TestandConfidenceIntervalTwosampleTforPKvsRPMNMeanStDevSEMeanPK202.12200.03870.0087RPM202.13950.02330.005295%CIformuPK-muRPM:(-0.0381,0.0031)T-TestmuPK=muRPM(vsnot=):T=-1.73P=0.093DF=31假设母分散不一样的附录1-14/41LGElectronics/LGENT6σTASKTEAM基础统计例题6:确认利用母平均的改善结果•运用英语分数的向上Program，比较Program进行前和进行后的英语分数，研讨向上Program在实际运用上是否有用.Program进行前/后的分数如下时，研讨Program对英语分数是否有用(各随机抽取10个样本)•Stat/basicstatistics/pairedtbeforeafter7681605285875870918675778290646379858883PairedT-TestandConfidenceIntervalPairedTforbefore-afterNMeanStDevSEMeanbefore1075.8011.643.68after1077.4012.183.85Difference10-1.606.382.0295%CIformeandifference:(-6.16,2.96)T-Testofmeandifference=0(vsnot=0):T-Value=-0.79P-Value=0.448-10010DifferencesBoxplotofDifferences(withHoand95%t-confidenceintervalforthemean)[]X_Ho附录1-15/41LGElectronics/LGENT6σTASKTEAM基础统计例题7:推定母比率p的置信区间•LGENT生产的显示器当中500个Sampling，调查结果判断为120个不良.求不良品比率p的95%近似置信区间•Stat/basicstatistics/1ProportionTestandConfidenceIntervalforOneProportionTestofp=0.5vspnot=0.5ExactSampleXNSamplep95.0%CIP-Value11205000.240000(