計量經(jīng)濟學(xué)試驗報告能源指數(shù)多重線性回歸

1、河用士林計量經(jīng)濟學(xué)實驗報告-經(jīng)濟貿(mào)易系國際經(jīng)濟與貿(mào)易專業(yè) 2014級2班41號實驗人王超實驗地點:實訓(xùn)樓 B305實驗日期:2016.10實驗題目：線性回歸模型實驗類型：基本操作實驗?zāi)康模?掌握Eviews的基本操作；散點圖的繪制；線性回歸模型中的參數(shù)估計、 模型檢驗、回歸預(yù)測的基本操作；并能夠針對回歸結(jié)果及檢驗結(jié)果做出正確的分析。 實驗內(nèi)容：（按要求完成下面題目）2.1表中是16支公益股票某年的每股帳面價值和當年紅利:公司序號帳面價值（元）紅利（元）公司序號帳面價值（元）紅利（元）122.442.4912.140.80220.892.981023.311.94322.092.061116.2

2、33.00414.481.09120.560.28520.731.96130.840.84619.251.551418.051.80720.372.161512.451.21826.431.601611.331.07根據(jù)上表資料：（1）繪制散點圖（2）建立每股帳面價值和當年紅利的回歸方程；（3）解釋回歸系數(shù)的經(jīng)濟意義；（4）對所建立的回歸模型進行檢驗。實驗步驟：2.1建立工作文彳建立 Workfile 后，在主界面命令欄鍵入 data y x,其中x代表賬面價值，y代表紅 利，錄入數(shù)據(jù)。(1)散點圖3.53.02.52.01.51.00.50.00481216202428估計參數(shù)Depende

3、nt Variable: YMethod: Least SquaresDate: 09/25/16 Time: 11:48Sample: 1 16Included observations: 16VariableCoefficientStd. Errort-StatisticProb.C0.4797750.3474401.3808830.1890X0.0728760.0194313.7505620.0022R-squared0.501189Mean dependent var1.671250Adjusted R-squared0.465559S.D. dependent var0.769813

4、S.E. of regression0.562775Akaike info criterion1.804596Sum squared resid4.434023Schwarz criterion1.901169Log likelihood-12.43676Hannan-Quinn criter.1.809541F-statistic14.06672Durbin-Watson stat1.976956Prob(F-statistic)0.002151Residual Actual FittedEstimation Command:=LS Y C XEstimation Equation: =Y

5、= C(1) + C(2)*XSubstituted Coefficients: =Y = 0.479774593451 + 0.0728758993263*X(2)回歸方程為 丫 = 0.479774593451 + 0.0728758993263*X(3)回歸系數(shù)的經(jīng)濟意義：所估計的參數(shù)BA1=0.4798 , BA2=0.072876 ,說明賬面價值每增加 1元，平均來說紅利增加了0.072826 ,這與預(yù)期的經(jīng)濟意義相符。（4）進行檢驗:Forecast: YFActual: YForecast sample: 1 16Included observations: 16Root Mea

6、n Squared Error0.526428Mean Absolute Error0.387470Mean Abs. Percent Error28.45368Theil Inequality Coefficient0.146944Bias Proportion0.000000Variance Proportion0.170997Covariance Proportion0.829003YF 2 S.E.統(tǒng)計檢驗1,擬合優(yōu)度：由表中數(shù)據(jù)可知RY=0.5011，修正的可決系數(shù)為0.4656,這說明模型對樣本的擬合不是很好。t檢驗（P值檢驗）由回歸結(jié)果可知 P=0.0022a=0.05,表明在a

7、=0.05的顯著性水平下每股賬面價值對每股紅利的影響是顯著的。2.2研究青春發(fā)育與遠視率（對數(shù)視力）的變化關(guān)系，測得結(jié)果如下表:年齡（歲）x遠視率（%） y對數(shù)視力丫=ln y663.644.153761.064.112838.843.659913.752.6211014.502.674118.072.088124.411.484132.270.82142.090.737151.020.02162.510.92173.121.138182.981.092試建立曲線回歸方程? = aebx （Y?= lna + bx）并進行計量分析。Dependent Variable: LNYMethod:

8、Least SquaresDate: 11/22/16 Time: 16:43Sample: 6 18Included observations: 13VariableCoefficientStd. Errort-StatisticProb.C5.7301980.6057009.4604630.0000X-0.3139400.048187-6.5150440.0000R-squared0.794184Mean dependent var1.962923Adjusted R-squared0.775473S.D. dependent var1.371926S.E. of regression0.

9、650076Akaike info criterion2.117184Sum squared resid4.648592Schwarz criterion2.204100Log likelihood-11.76170Hannan-Quinn criter.2.099319F-statistic42.44580Durbin-Watson stat0.662056Prob(F-statistic)0.000043lna=5.7302 b=-0.3139，所以 y=lna+bx=5.7302-0.3139x從回歸結(jié)果來看，RA2=0.7942 ,所以模型擬合一般。RA2=0.7942 ,t值為9.4

10、605-6.5150 ,斜率項為-0.3139, 這表明孩子的對數(shù)視力隨著年齡的增加呈平均下降趨勢。3.1下表給出的是1960 1982年間7個OECD國家的能源需求指數(shù)（丫）、實際GDP指數(shù)（X1）、能源價格指數(shù)（X2）的數(shù)據(jù)，所有指數(shù)均以1970年為基準（ 1970=100）年份能源需求指數(shù)Y實際GDP指數(shù)X1能源價格指數(shù)X2年份能源需求指數(shù)Y實際GDP指數(shù)X1能源價格指數(shù)X2196054.154.1111.9197297.294.398.6196155.456.4112.41973100.0100.0100.0196258.559.4111.1197497.3101.4120.11963

11、61.762.1110.2197593.5100.5131.0196463.665.9109.0197699.1105.3129.6196566.869.5108.31977100.9109.9137.7196670.373.2105.31978103.9114.4133.7196773.575.7105.41979106.9118.3144.5196878.379.9104.31980101.2119.6179.0196983.383.8101.7198198.1121.1189.4197088.986.297.7198295.6120.6190.9197191.889.8100.3(1)建

12、立能源需求與收入和價格之間的對數(shù)需求函數(shù)lnY = Po + A In X1t + 02 1nx2 +5 ,解釋各回歸系數(shù)的意義，用P值檢驗所估計回歸系數(shù)是否顯著。(2)再建立能源需求與收入和價格之間的線性回歸模型Yt = Po + PiXIt + p2X2t + u ,解釋各回歸系數(shù)的意義，用P值檢驗所估計回歸系數(shù)是否顯著。(3)比較兩個模型參數(shù)估計結(jié)果的經(jīng)濟意義有什么不同？(4)如果兩個模型結(jié)論不同，你將選擇哪個模型，為什么？Dependent Variable: LNYMethod: Least SquaresDate: 11/22/16 Time: 21:44Sample: 1 23I

13、ncluded observations: 23VariableCoefficientStd. Errort-StatisticProb.C1.5495040.09011317.195090.0000LNX10.9969230.01911052.166360.0000LNX2-0.3313640.024310-13.630870.0000R-squared0.994130Mean dependent var4.412077Adjusted R-squared0.993543S.D. dependent var0.224107S.E. of regression0.018008Akaike in

14、fo criterion-5.074917Sum squared resid0.006486Schwarz criterion-4.926809Log likelihood61.36154Hannan-Quinn criter.-5.037668F-statistic1693.654Durbin-Watson stat0.807847Prob(F-statistic)0.000000實驗結(jié)果：對模型的參數(shù)進行估計，根據(jù)回歸結(jié)果得：LnAYt=1.549504+0.996923lnX1t-0.331364lnX2tt=(17.19508) (52.16634) (-13.63086)RA2=0.

15、99413 - RA2=0.993543 F=1693.652S.E.=0.018008 DW=0.807846經(jīng)濟意義：在能源價格指數(shù)不變的情況下，實際GDP旨數(shù)的對數(shù)每增加一個單位，能源需求指數(shù)的對數(shù)平均增加0.996923個單位；在實際GD指數(shù)不變的情況下，能源價格指數(shù)的對數(shù)每增加一個單位，能源需求指數(shù)的對數(shù)平均減少0.331364個單位。B A1, 3人2的P值均為0.0000 ,遠遠小于0.05 ,說明回歸系數(shù)均顯著。實驗步驟：Dependent Variable: YMethod: Least SquaresDate: 11/22/16 Time: 21:50Sample: 1 2

16、3Included observations: 23VariableCoefficientStd. Errort-StatisticProb.C28.255061.42148819.877090.0000X10.9808490.01945450.419000.0000X2-0.2584260.015282-16.910310.0000R-squared0.993890Mean dependent var84.34348Adjusted R-squared0.993279S.D. dependent var17.50999S.E. of regression1.435479Akaike info

17、 criterion3.681982Sum squared resid41.21199Schwarz criterion3.830090Log likelihood-39.34279Hannan-Quinn criter.3.719230F-statistic1626.707Durbin-Watson stat0.977840Prob(F-statistic)0.000000實驗結(jié)果：對模型的參數(shù)進行估計，根據(jù)回歸結(jié)果得：LnAYt=28.25506+0.980849lnX1t-0.30.258426lnX2tt=(19.87709) (50.419) (-16.91031)RA2=0.99389 - RA2=0.993279 F=1626.707S.E.=1.435479 DW=0.977840經(jīng)濟意義：在能源價格指數(shù)不變的情況下，實際GDP旨數(shù)的對數(shù)每增加一個單位，能源需求指數(shù)的對數(shù)平均增加0.9808049個單位；在實際GDP旨數(shù)不變的情況下，能源價格指數(shù)的對數(shù)每增加