您好,欢迎访问三七文档
当前位置:首页 > 商业/管理/HR > 招聘面试 > 温州医科大学研究生统计复习
一:单样本T检验:例3-5某医生测量了36名从事铅作业男性工人的血红蛋白含量,算得其均数为130.83g/L,标准差为25.74g/L。问从事铅作业工人的血红蛋白是否不同于正常成年男性平均值140g/L?One-SampleTest单样本检验TestValue实验值=140tDf自由度Sig.(2-tailed)P值(双侧)MeanDifference均数差值95%ConfidenceIntervaloftheDifference差值的95%可信区间LowerUpper血红蛋白含量-2.13735.040-9.16667-17.8762-.4572二:配对样本T检验:例3-6为比较两种方法对乳酸饮料中脂肪含量测定结果是否不同,随机抽取了10份乳酸饮料制品,分别用脂肪酸水解法和哥特里-罗紫法测定其结果如表3-5第(1)~(3)栏。问两法测定结果是否不同?表3-5两种方法对乳酸饮料中脂肪含量的测定结果(%)编号(1)哥特里-罗紫法(2)脂肪酸水解法(3)差值d(4)=(2)(3)10.8400.580错误!未找到引用源。20.5910.509错误!未找到引用源。30.6740.500错误!未找到引用源。40.6320.316错误!未找到引用源。50.6870.337错误!未找到引用源。60.9780.517错误!未找到引用源。70.7500.454错误!未找到引用源。80.7300.512错误!未找到引用源。91.2000.997错误!未找到引用源。100.8700.506错误!未找到引用源。错误!未找到引用源。PairedSamplesTest配对样本检验PairedDifferencestdf自由度Sig.(2-tailed)双侧P值Mean差值均数Std.Deviation标准差Std.ErrorMean标准误95%ConfidenceIntervaloftheDifference差值的95%可信区间LowerUpperPair1哥特里-罗紫法脂肪酸水解法.272400.108681.034368.194654.3501467.9269.000三:两样本T检验:(成组设计两样本T检验)例3-7为研究国产四类新药阿卡波糖胶囊的降血糖效果,某医院用40名2型糖尿病病人进行同期随机对照试验。试验者将这些病人随机等分到试验组(用阿卡波糖胶囊)和对照组(用拜糖平胶囊),分别测得试验开始前和8周后的空腹血糖,算得空腹血糖下降值见表3-6,能否认为该国产四类新药阿卡波糖胶囊与拜糖平胶囊对空腹血糖的降糖效果不同?表3-6试验组和对照组空腹血糖下降值(mmol/L)试验组X1(n1=20)-0.70-5.602.002.800.703.504.005.807.10-0.502.50-1.601.703.000.404.504.602.506.00-1.40对照组X2(n2=20)3.706.505.005.200.800.200.603.406.60-1.106.003.802.001.602.002.201.203.101.70-2.00IndependentSamplesTesT独立样本T检验Levene'sTestforEqualityofVariancesF方差齐性检验t-testforEqualityofMeans两均数是否相等的T检验FSig.tDf自由度Sig.(2-tailed)双侧P值MeanDifference均数差值Std.ErrorDifference差值标准误95%ConfidenceIntervaloftheDifference差值的95%可信区间LowerUpper空腹血糖下降值Equalvariancesassumed假设方差齐性.578.452-.64238.525-.56000.87245-2.326181.20618Equalvariancesnotassumed假设方差不齐-.64236.086.525-.56000.87245-2.329261.209264完全随机设计资料的方差分析:例4-2某医生为了研究一种降血脂新药的临床疗效,按统一纳入标准选择120名高血脂患者,采用完全随机设计方法将患者等分为4组(具体分组方法见例4-1),进行双盲试验。6周后测得低密度脂蛋白作为试验结果,见表4-3。问4个处理组患者的低密度脂蛋白含量总体均数有无差别?表4-34个处理组低密度脂蛋白测量值(mmol/L)分组测量值统计量niXX2X安慰剂组3.534.594.342.663.593.132.642.563.503.25303.43102.91367.853.304.043.533.563.854.073.523.934.192.961.373.932.332.984.003.552.964.34.162.59降血脂新药302.7281.46233.002.4g组2.423.364.322.342.682.951.563.111.811.771.982.632.862.932.172.722.652.222.902.972.362.562.522.272.983.722.803.574.022.314.8g组2.862.282.392.282.482.283.212.232.322.68302.7080.94225.542.662.322.613.642.583.652.663.682.653.023.482.422.412.663.292.703.042.811.971.687.2g组0.891.061.081.271.631.891.192.172.281.72301.9758.99132.131.981.742.163.372.971.690.942.112.812.521.312.511.881.413.191.922.471.022.103.71ANOVA方差分析SumofSquares离均差平方和Df自由度MeanSquare均方MSFSig.P值BetweenGroups组间变异32.156310.71924.884.000WithinGroups组内变异49.967116.431Total总变异82.1231195随机区组设计资料的方差分析:例4-4某研究者采用随机区组设计进行实验,比较三种抗癌药物对小白鼠肉瘤抑瘤效果,先将15只染有肉瘤小白鼠按体重大小配成5个区组,每个区组内3只小白鼠随机接受三种抗癌药物(具体分配方法见例4-3),以肉瘤的重量为指标,试验结果见表4-9。问三种不同的药物的抑瘤效果有无差别?表4-9不同药物作用后小白鼠肉瘤重量(g)区组A药B药C药1gijiX10.820.650.511.9820.730.540.231.5030.430.340.281.0540.410.210.310.9350.680.430.241.351nijjX3.072.171.576.81()ijXiX0.6140.4340.3140.454()X21nijjX2.02071.05870.54513.62452()ijXTestsofBetween-SubjectsEffectsDependentVariable:weightSource变异来源TypeIIISumofSquares三型离均差平方和Df自由度MeanSquare均方MSF统计量FSig.P值CorrectedModel校正的模型.456(a)6.0767.964.005Intercept截距3.09213.092323.742.000Group区组间.2284.0575.978.016Treat处理间.2282.11411.937.004Error误差.0768.010Total合计3.62415CorrectedTotal总变异(校正的合计).53314aRSquared=.857(AdjustedRSquared=.749)6拉丁方设计资料的方差分析:例4-5某研究者为了比较甲、乙、丙、丁、戊、己6种药物给家兔注射后产生的皮肤疱疹大小(mm2),采用拉丁方设计,选用6只家兔、并在每只家兔的6个不同部位进行注射。试验结果见表4-11,试作方差分析。表4-11例4-5拉丁方设计与试验结果(皮肤疱疹大小,mm2)家兔编号(行区组)注射部位编号(列区组)行区组合计(Rj)iX123456TestsofBetween-SubjectsEffectsDependentVariable:实验结果Source变异来源TypeIIISumofSquares三型离均差平方和df自由度MeanSquare均方MSF统计量FSig.P值CorrectedModel校正的模型982.948(a)1565.5301.863.096Intercept截距198366.0771198366.0775640.545.000行区组(家兔间)251.663550.3331.431.256列区组(部位间)65.337513.067.372.862处理组(药物间)657.3365131.4673.738.015Error误差703.3582035.168Total合计202241.00036CorrectedTotal总变异(校正的合计)1686.30635aRSquared=.583(AdjustedRSquared=.270)7两阶段交叉设计资料的方差分析:例4-6表4-13是A、B两种闪烁液测定血浆中3H-cGMP的交叉试验结果。第I阶段1、3、4、7、9号用A测定,2、5、6、8、10号用B测定;第II阶段1、3、4、7、9号用B测定,2、5、6、8、10号用A测定。试对交叉试验结果进行方差分析。表4-13两种闪烁液测定血浆中3H-cGMP的交叉试验受试者阶段合计BiⅠⅡ1A(760)B(770)15302B(860)A(855)17153A(568)B(602)11704A(780)B(800)15805B(960)A(958)19186B(940)A(952)18927A(635)B(650)12858B(440)A(450)8909A(528)B(530)105810B(800)A(803)1603各阶段合计A、B合计S1=7271T1=7289S2=7370T2=735214641(∑X)TestsofBetween-SubjectsEffectsDependentVariable:Ѫ½¬3H-cGMP1C(87)B(75)E(81)D(75)A(84)F(66)468.078.02B(73)A(81)D(87)C(85)F(64)E(79)469.078.23F(73)E(73)B(74)A(78)D(73)C(77)448.074.74A(77)F(68)C(69)B(74)E(76)D(73)437.072.85D(64)C(64)F(72)E(76)B(70)A(81)427.071.26E(75)D(77)A(82)F(61)C(82)B(61)438.073.0列区组合计(Ci)449.0438.0465.0449.0449.0437.0jX74.873.077.574.874.872.8药物DECABFX=74.6合计Tk449.0483.0464.0483.0427.0404.0kX74.880.577.380.571.267.3Source变异来源TypeIIISumofSquares三型离均差平方和df自由度MeanSquare均方MSF统计量FSig.P值CorrectedModel校正的模型551799.950(a)1150163.6321015.972.000Intercept截距10717944.050110717944.050217072.285.000Person受试者间551111.450961234.6061240.195.000Treat处理间(A。B)198.4501198.4504.019.080Phase阶段间(1,2)490.0501490.0509.925.01
本文标题:温州医科大学研究生统计复习
链接地址:https://www.777doc.com/doc-2244129 .html