The results of the regression with EPI as the dependent variable
are presented in Table 2. The significant variables that affect EPI
are operating hours per week, logarithm of the area, employment
density and the presence of a VFD. These four variables together
explain 70.6% of the variance in EPI (adj R-squared = 0.706). The
Variance Inflation Factors (VIF) for each variable is less than two
and shows no multicollinearity effects.