The RATS Software Forum

Dear Tom,

I am some how confused about the below results.
I would be grateful if you could possibly guide me.

How it is possible that:
a) the null of beta=0 cannot be rejected and meanwhile the null of beta=1 also cannot be rejected?

And what about the the joint hypothesis of Alpha=0 and Beta=1? (it cannot be rejected).

Linear Regression - Estimation by Least Squares
With Heteroscedasticity-Consistent (Eicker-White) Standard Errors
Dependent Variable DEP
Usable Observations                        53
Degrees of Freedom                         51
Centered R^2                        0.0200152
R-Bar^2                             0.0007998
Uncentered R^2                      0.1113663
Mean of Dependent Variable       0.0371020898
Std Error of Dependent Variable  0.1168258428
Standard Error of Estimate       0.1167791170
Sum of Squared Residuals         0.6955054710
Log Likelihood                        39.6316
Durbin-Watson Statistic                2.2798

    Variable                        Coeff      Std Error      T-Stat      Signif
************************************************************************************
1.  Constant                     0.0169966283 0.0214227068      0.79339  0.42754881
2.  INDEP                        0.6303479148 0.9810692176      0.64251  0.52054135


Test for indep(Beta)=1
Chi-Squared(1)=      0.141967 with Significance Level 0.70633345


Test for constant(Alpha)=0 & indep(Beta)=1
Chi-Squared(2)=      1.137291 or F(2,*)=      0.56865 with Significance Level 0.56629188

Your interpretation is correct; you cannot reject either 0 or 1. Obviously the data don't have very strong information about the value of beta.