Do I need to rescale variables to compare coefficients.

[A_Salem]

I'm running a regression analysis. My "Y" variable is Yearly_Spend while my "X" variables are Time_Spent (on website in minutes) and Length (of membership in years). Right now my results show Years has a bigger impact on Yearly_Spend. Do get a true apples to apples comparison of the coefficients, does it make sense to rescale my variables and run the analysis again?

 

[hlsmith]

If they were standardized then you would have a comparison in regards to a standard deviation increase in either variable, so yes that is what would be done traditionally. You can also report their partial r^2 values.