经济预测与策计算题复习题.doc

《经济预测与策计算题复习题.doc》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、P596、SUMMARYOUTPUT回归统计MultipleR0.9976RSquare0.9953AdjustedRSquare0.9945标准误差0.4965观测值8.0000方差分析　dfSSMSFSignificanceF回归分析1.0000311.3960311.39601263.23380.0000残差6.00001.47900.2465总计7.0000312.8750　　　　Coefficients标准误差tStatP-valueLower95%Upper95%下限95.0%上限95.0%Intercept2.08980.40665.13980.002

2、11.09493.08471.09493.0847XVariable11.93110.054335.54200.00001.79822.06411.79822.0641RESIDUALOUTPUT观测值预测Y残差(e-(e-1))^2e^21.00005.95210.04790.00232.00007.88320.11680.00470.01363.000011.7455-0.74550.74350.55584.000013.67660.32341.14250.10465.000015.60780.39220.00470.15386.000019.4701-0.47

3、010.74350.22107.000021.40120.59881.14250.35868.000025.2635-0.26350.74350.06944.52501.4790DW==3.0594DependentVariable:YMethod:LeastSquaresDate:10/24/13Time:10:52Sample:18Includedobservations:8VariableCoefficientStd.Errort-StatisticProb.  C2.0898200.4065985.1397690.0021X1.9311380.0543343

4、5.542000.0000R-squared0.995273    Meandependentvar15.12500AdjustedR-squared0.994485    S.D.dependentvar6.685539S.E.ofregression0.496495    Akaikeinfocriterion1.649830Sumsquaredresid1.479042    Schwarzcriterion1.669690Loglikelihood-4.599320    Hannan-Quinncriter.1.515880F-statistic1263.

5、234    Durbin-Watsonstat3.059395Prob(F-statistic)0.000000从上述数据可知该模型的回归方程是：Y=2.0898+1.9311x拟合优度：R^2=0.9953表明Y的变异被X的变异解释了99.53%，说明X与Y的拟合优度非常好。F检验：假设=0；：0F=1263.2338F=1263.2338>(1,6)=5.99所以否定原假设，说明回归方程非常显著。T检验：假设=0；：0T=35.5420=35.5420>(6)=2.447所以参数估计值都能通过t检验，在统计上是显著的，可以认为广告费用支出对销售额有显著性影响。

6、DW检验：DW=3.0594(n=8,k=1)由DW统计表可以看到，当n=15，自变量个数k=1,=1.08,=1.36,能够确定4-DW4所以可以判断模型随机项存在序列负自相关。点预测:当X=16万元时，Y=32.9874万元根据线性预测模型，我们可以预测，当广告费用为16万元时，且当显著性水平为5%时，该该公司下一季度的销售额是32.9874万元。区间预测：S=0.4965=0.4965*1.466=0.727932.9874-1.96*0.7279=31.558632.9874+1.96*0.7279=34.4140即置信度为95%的预测区间为（31.56,3

7、4.42）7、SUMMARYOUTPUT回归统计MultipleR0.937723765RSquare0.87932586AdjustedRSquare0.84484753标准误差1.966842176观测值10方差分析　dfSSMSFSignificanceF回归分析2197.32072398.6603614925.5037290.00061045残差727.0792773.868468146总计9224.4　　　　Coefficients标准误差tStatP-valueLower95%Upper95%下限95.0%上限95.0%Intercept4.5875