Multiple Regression

My data set contains five variables for 96 nations in the world.

 

Onlinepop        Online Population

PC's                 Number of Personal Computeres

Phones            Number of landline phones

Educ                Percent of GNP spent on education

GNPPC            Gross National Product per capita

 

Here is correlation matrix on data set:

 

  Onlinepop PCs Phones Educ GNPPC

Onlinepop 1





PCs 0.990643 1



Phones 0.319927 0.275276 1



Educ 0.049997 0.049423 0.369801 1

GNPPC 0.509078 0.477851 0.874735 0.318304 1

 

 

 

 

 

 

 

 

Assigment was to Regress Onlinepop against PC's, Phones, and Educ and
then to regress the predicted values of the dependent variable Onlinepop
against the residuals resulting in the following scatterplot in order
to detect for heteroskedasticity.

Next based on results of model I was to consider three scenarios. 
First, triple Education expenditure.  Second, double PC's.  Third,
double Phones.  Here are results of model.  R square  was .98

 

                               
Coefficients        Standard Error            T
stat        P-value

Intercept                     544579.306   
    613179.8725            0.89            .38

PCs                        0.937418021   
     0.013151766           
71.28          2.67337E-82

Phones                   0.218685792         
0.054724064            3.99           .0001

Educ                     -187381.0927       
   132987.9076           -1.40            .1622

 

QUESTION - I don't understand why variable Educ has a negative
coefficient.  Intutitively, I would expect a positive sign.  Can you
explain why?  Could it be multilcollinearity or is it the
heteroskedasticity at work?

 

Thanks,

 

Robin Otto