University of Westminster

Emperical Analysis  

 

Before moving to Emperical analysis we convert our dependent variable and some of independent variables to 

“Ln” form. To evaluate systematic risk of capital market stock’s return Leigh McAlister, Raji Srinivasan, & 

MinChung Kim (2007) used ln(D

it

+P

it

) and ln(L

t

/L

t-1

) where D

it 

stands for cash dividend payable, P

it 

is the 

price of ordinary share. L

t

/L

t-1 

is index of all NYSE stock price. So to clearly interpretative the result and 

taking into consideration large number in observations from table 1 it is more appropriate to convert “Total 

annual sales” (dependent variable), “Spending on purchases of Fixed assets” and “Expenditures on R&D within 

 

      When degree of correlation between independent variables is high there might be a problem of providing 

unique and independent information in the regression model. In order to detect we first used a metric called VIF 

(the variance inflation factor), in this metric we measured strength of correlation between the independent 

variables in our model.