高手们帮帮我呀，我最近要写一篇毕业论文，在网上看到例文上面有人用eviews分析股票收益率问题。就下载了一个按照他的步骤分析了一下，数据都代对了，但是结果里的prob.和F-statistics不一样。我算的结果如下：


Dependent Variable: GPI   
Method: Least Squares   
Date: 03/13/10   Time: 23:58   
Sample: 1991 2007   
Included observations: 17   
Estimation settings: tol= 0.00010, derivs=analytic (linear)   
GPI=C(1)+C(2)*M+C(3)*ROU+C(4)*FD+C(5)*R+C(6)*GDP   
   
Coefficient Std. Error t-Statistic Prob.  
   
C(1) 2351.988 1988.990 1.182504 0.2619
C(2) 0.004596 0.016075 0.285884 0.7803
C(3) -486.2785 983.6558 -0.494358 0.6308
C(4) 0.225229 0.594603 0.378789 0.7121
C(5) -75.03728 122.1036 -0.614538 0.5514
C(6) 0.007362 0.030979 0.237653 0.8165
   
R-squared 0.634997     Mean dependent var  1419.610
Adjusted R-squared 0.469086     S.D. dependent var  906.2865
S.E. of regression 660.3551     Akaike info criterion  16.09400
Sum squared resid 4796757.     Schwarz criterion  16.38807
Log likelihood -130.7990     Durbin-Watson stat  1.312377
   
论文里的结果是这样的

Dependent Variable: GPI   
Method: Least Squares   
Sample: 1991 2007   
Included observations: 17   
   
Coefficient Std. Error t-Statistic Prob.  
   
C(1) 2351.988 1988.990 1.182504 0.0619
C(2) 0.004596 0.016075 0.285884 0.0803
C(3) -486.2786 983.6558 -0.494358 0.0308
C(4) 0.225229 0.594603 0.378789 0.0121
C(5) -75.03728 122.1036 -0.614538 0.0514
C(6) 0.007362 0.030979 0.237653 0.0165
   
R-squared 0.934997     Mean dependent var  1419.610
Adjusted R-squared 0.969086     S.D. dependent var  906.2865
S.E. of regression 660.3551     Akaike info criterion  16.09400
Sum squared resid 4796757.     Schwarz criterion  16.38807
Log likelihood -130.7990     Durbin-Watson stat  1.312377
F-statistics  3.827343        prob(f-statistics)   0.000672




数据都没有错误，是步骤错了还是怎样望各位大大帮下忙。万分感谢

btw：我这里有论文的原本和数据，QQ：124991660