实验练习题1、根据美国各航空公司航班正点到达的比率X(%)和每10万名乘客投诉的次数Y进行回归,EViews输出结果如下:DependentVariable:YMethod:LeastSquaresSample:19Includedobservations:9VariableCoefficientStd.Errort-StatisticProb.C6.0178321.0522605.7189610.0007X-0.0704140.014176-4.9672540.0016R-squared0.778996Prob(F-statistic)0.001624Durbin-Watsonstat2.5270(1)对以上结果进行简要分析(包括方程显著性检验、参数显著性检验、DW值的评价、对斜率的解释等,显著性水平均取0.05)。(2)按标准书写格式写出回归结果。2、已知变量Y和X的数据如下表所示,试采用OLS法(列出表格)估计模型iY=01iiXu的参数值。序号113222338448551166133、以下是某次线性回归的EViews输出结果,部分数值已略去(用大写字母标示),但它们和表中其它特定数值有必然联系,分别据此求出这些数值,并写出过程。(保留3位小数)DependentVariable:YMethod:LeastSquaresSample:113Includedobservations:13VariableCoefficientStd.Errort-StatisticProb.C5.7304880.605747A0.0000X-0.3139600.048191-6.5149640.0000R-squared0.794180Meandependentvar1.962965AdjustedR-squaredBS.D.dependentvar1.372019S.E.ofregression0.650127Akaikeinfocriterion2.117340SumsquaredresidCSchwarzcriterion2.2042564、用1970-1994年间日本工薪家庭实际消费支出Y与实际可支配收入X(单位:103日元)数据估计线性模型Y=01Xu,然后用得到的残差序列te绘制以下图形。(1)试根据图形分析随机误差项之间是否存在自相关?若存在,是正自相关还是负自相关?(2)此模型的估计结果为t:(6.14)(30.01)2R=0.975,F=900.51,DW=0.35试用DW检验法检验随机误差项之间是否存在自相关。5、用一组截面数据估计消费(Y)—收入(X)方程Y=01Xu的结果为iY=9.3480.637iXt:(2.57)(32.01)2R=0.95,F=1024.56,DW=1.79(1)根据回归的残差序列e(t)图分析本模型是否存在异方差?注:abs[e(t)]表示e(t)的绝对值。(2)其次,用White法进行检验。EViews输出结果见下表:WhiteHeteroskedasticityTest:F-statistic6.301373Probability0.003370Obs*R-squared10.86401Probability0.004374DependentVariable:RESID^2Method:LeastSquaresSample:160Includedobservations:60VariableCoefficientStd.Errort-StatisticProb.C-10.03614131.1424-6.0765290.0045X0.1659771.6198565.1024640.0064X^20.0018000.0045878.3924690.0002若给定显著水平0.05,以上结果能否说明该模型存在异方差?查卡方分布临界值的自由度是多少?6.下表是中国某地人均可支配收入(INCOME)与储蓄(SAVE)之间的回归分析结果(单位:元):DependentVariable:SAVEMethod:LeastSquaresSample:131Includedobservations:31VariableCoefficientStd.Errort-StatisticProb.C-695.1433118.0444-5.8888270.0000INCOME0.0877740.004893――――R-squared0.917336Meandependentvar1266.452AdjustedR-squared0.914485S.D.dependentvar846.7570S.E.ofregression247.6160Akaikeinfocriterion13.92398Sumsquaredresid1778097.Schwarzcriterion14.01649Loglikelihood-213.8216F-statistic321.8177Durbin-Watsonstat1.892420Prob(F-statistic)0.000000附表:DW检验临界值表(=0.05)nk=1k=2dLdUdLdU241.271.451.191.55251.291.451.211.55261.301.461.221.55271.311.471.241.561)请写出样本回归方程表达式,然后分析自变量回归系数的经济含义2)解释样本可决系数的含义3)写出t检验的含义和步骤,并在5%的显著性水平下对自变量的回归系数进行t检验(临界值:t0.025(29)=2.05)。4)下表给出了White异方差检验结果,试在5%的显著性水平下判断随机误差项是否存在异方差。WhiteHeteroskedasticityTest:F-statistic6.048005Probability0.006558Obs*R-squared9.351960Probability0.0093165)下表给出LM序列相关检验结果(滞后1期),试在5%的显著性水平下判断随机误差项是否存在一阶自相关。Breusch-GodfreySerialCorrelationLMTest:F-statistic0.030516Probability0.862582Obs*R-squared0.0337...