统计学上机作业.doc

统计软件应用作业（3）方差分析作业公管11 2111401025 潘烨烽 1、抽查某地区三所小学五年级男学生的身高，数据见文件：“男生身高”。设各小学五年级男学生的身高服从同方差的正态分布。问该地区三所小学五年级男学生的平均身高是否有显著差异（=0.05）？ ANOVA身高Sum of SquaresdfMean SquareFSig.Between Groups465.8812232.9414.372.032Within Groups799.2551553.284Total1265.13617Multiple Comparisons身高LSD(I) 编号(J) 编号Mean Difference (I-J)Std. ErrorSig.95% Confidence IntervalLower BoundUpper Bound12-10.8500*4.2144.021-19.833-1.8673-10.7333*4.2144.022-19.716-1.7512110.8500*4.2144.0211.86719.8333.11674.2144.978-8.8669.0993110.7333*4.2144.0221.75119.7162-.11674.2144.978-9.0998.866*. The mean difference is significant at the 0.05 level.P=.032.05，有显著差异2、某钢厂检查一月上旬内的五天中生产钢锭重量，数据见文件：“钢锭重量”，设各日所生产的钢锭重量服从同方差的正态分布，试检验不同日期生产的钢锭的平均重量有无显著差异（=0.05）？ANOVA重量Sum of SquaresdfMean SquareFSig.Between Groups227680.000456920.0003.950.022Within Groups216175.0001514411.667Total443855.00019Multiple Comparisons重量LSD(I) 日期(J) 日期Mean Difference (I-J)Std. ErrorSig.95% Confidence IntervalLower BoundUpper Bound12197.5000*84.8872.03416.567378.4334277.5000*84.8872.00596.567458.433912.500084.8872.885-168.433193.43310107.500084.8872.225-73.433288.43321-197.5000*84.8872.034-378.433-16.567480.000084.8872.361-100.933260.9339-185.0000*84.8872.046-365.933-4.06710-90.000084.8872.306-270.93390.93341-277.5000*84.8872.005-458.433-96.5672-80.000084.8872.361-260.933100.9339-265.0000*84.8872.007-445.933-84.06710-170.000084.8872.064-350.93310.93391-12.500084.8872.885-193.433168.4332185.0000*84.8872.0464.067365.9334265.0000*84.8872.00784.067445.9331095.000084.8872.281-85.933275.933101-107.500084.8872.225-288.43373.433290.000084.8872.306-90.933270.9334170.000084.8872.064-10.933350.9339-95.000084.8872.281-275.93385.933*. The mean difference is significant at the 0.05 level.Sig=0.0340.05，所以样本方差不齐P=.0220.05，所以有显著性差异3、在某种橡胶的配方中，考虑了3种不同的促进剂，4种不同分量的氧化剂。各种配方各实验一次，测得300%定强数据见文件：“橡胶配方定强”。假定各种配方的定强服从同方差的正态分布。试问不同促进剂、不同分量氧化锌分别对定强有无显著影响（=0.05）？Between-Subjects FactorsValue LabelN促进剂1促进剂A142促进剂A243促进剂A34氧化锌1氧化锌B132氧化锌B233氧化锌B334氧化锌B43Tests of Between-Subjects EffectsDependent Variable:配方定强SourceType III Sum of SquaresdfMean SquareFSig.Corrected Model94.354a518.87147.674.000Intercept16465.021116465.02141595.842.000促进剂28.292214.14635.737.000氧化锌66.062322.02155.632.000Error2.3756.396Total16561.75012Corrected Total96.72911a. R Squared = .975 (Adjusted R Squared = .955)表中促进剂，氧化锌的sig=.0000.05，有显著性差异4、某灯泡厂用四种不同配料方案制成的灯丝，生产了四批灯泡。在每批灯泡中随机的抽取若干个灯泡测其使用寿命（单位：小时），数据见文件：“灯泡使用寿命”。试问对于这四种灯丝生产的灯泡，其使用寿命有无显著差异（=0.05）。绘制均值散布图，说明那种配料方案最佳。ANOVA寿命Sum of SquaresdfMean SquareFSig.Between Groups39776.456313258.8191.638.209Within Groups178088.929228094.951Total217865.38525Multiple Comparisons寿命LSD(I) 灯丝(J) 灯丝Mean Difference (I-J)Std. ErrorSig.95% Confidence IntervalLower BoundUpper Bound1276.28652.682.162-32.97185.54325.53646.565.589-71.03122.11499.28650.056.060-4.52203.1021-76.28652.682.162-185.5432.973-50.75051.292.333-157.1255.62423.00054.481.677-89.99135.9931-25.53646.565.589-122.1171.03250.75051.292.333-55.62157.12473.75048.590.143-27.02174.5241-99.28650.056.060-203.104.522-23.00054.481.677-135.9989.993-73.75048.590.143-174.5227.02表1，sig=0.0610.05，所以样本方差齐表2，p=0.2090.05，所以对四种灯丝生产的灯泡，使用寿命无显著差异由图可知，配料方案1最佳5、用四种饲料喂猪，共19头猪分成四组，每组用一种饲料，一段时间后称重。猪体重增加数据见文件DATA12-01，试比较这四种饲料对猪体重增加的作用有无显著差异（=0.05）。通过均值多重比较，说明那种饲料最合适。ANOVAweightSum of SquaresdfMean SquareFSig.Between Groups20538.69836846.233157.467.000Within Groups652.1591543.477Total21190.85818Multiple ComparisonsweightLSD(I) fodder(J) fodderMean Difference (I-J)Std. ErrorSig.95% Confidence IntervalLower BoundUpper BoundAB-18.68000*4.17024.000-27.5687-9.7913C-56.36000*4.17024.000-65.2487-47.4713D-87.41500*4.42321.000-96.8428-77.9872BA18.68000*4.17024.0009.791327.5687C-37.68000*4.17024.000-46.5687-28.7913D-68.73500*4.42321.000-78.1628-59.3072CA56.36000*4.17024.00047.471365.2487B37.68000*4.17024.00028.791346.5687D-31.05500*4.42321.000-40.4828-21.6272DA87.41500*4.42321.00077.987296.8428B68.73500*4.42321.00059.307278.1628C31.05500*4.42321.00021.627240.4828*. The mean difference is significant at the 0.05 level.表1，sig=0.9950.05，所以样本方差齐表2，p=0.0000.05，则四种饲料对猪体重增加的作用有显著差异由图，饲料D最合适6、研究两种治疗缺铁性贫血药物的疗效，随机抽取12个病人分成四组，给以不同的疗效：第一组使用一般疗法；第二组使用一般疗法外加药物A；第三组使用一般疗法外加药物B；第四组在一般疗法外家用药物A药物B。一个月后观察红细胞增加数（百万/mm3），数据见文件：DATA12-04。请问药物A,药物B是否分别对红细胞增加数有显著疗效？同时使用药物A和药物B是否对红细胞增加数有显著疗效（=0.05）？绘制均值散布图，说明那种治疗方案最佳。Between-Subjects FactorsValue LabelN药A0不用61用6药B0不用61用6Tests of Between-Subjects EffectsDependent Variable:红细胞增数SourceType III Sum of SquaresdfMean SquareFSig.Corrected Model2.963a3.98898.750.000Intercept19.508119.5081950.750.000druga1.68811.688168.750.000drugb.9081.90890.750.000druga * drugb.3681.36836.750.000Error.0808.010Total22.55012Corrected Total3.04311a. R Squared = .974 (Adjusted R Squared = .964)表1，sig=1.0000.05，所以样本方差齐表2，p=0.0000.05，所以药物A，药物B分别对细胞增加数有显著疗效，同时使用药物A和药物B对细胞增加数有显著疗效，说明药物A和药物B有交互作用由图可知，一般疗法外加药物A和药物B的治疗方案最佳