How can I perform a statistical test to judge impact of an independent variable on a dependent variable given multiple regression output? In the example below, how would I test if gender or work experience has a greater impact on gross pay?
Independent variables: education, work experience Dependent variable: gross pay
Example (artificial) Multiple regression output:
Predictor Coefficient SECoef T
Constant 1200 .35 3.43
Education 3500 .020 1.75
Work exp. 3000 .012 2.4
I understand that looking at the coefficients should be the first place we can look to determine the affect ind. variables have on the dependent variable. In this case, education would have more of an effect on the dependent variable because its coefficient is larger than the coefficient for work experience. I also realize that the T value of work experience is larger, indicating it is more statistically significant than education.
Is there a better way to gauge which ind. variable has a larger affect on the dependent variable? Perhaps with a CI/hypothesis test?
