Hi, as part of my work I'm running panel data linear regression on 2 variables: the independent variable is also a binary variable (it's values are: 0/1), and the the dependent variable is a continues LOG variable (LN action was set on it, its values are 0-20).
My function I'm using is: xtreg dependent_log_variable independent_binary_variable , re (also for fixed effects - xtreg dependent_log_variable independent_binary_variable , fe)
I was told that in this case I shouldn't get as result of the coefficient (beta) to be larger than 1, but I got 1.83. My data values seems to be ok.
My question is:
- is my beta coefficient of 1.83 sounds reasonable to get? if not - why shouldn't I get such result in this case?
- just to verify: in order to get the real value of the coefficient I'm using the follow equation: ( (exp(beta) - 1) *100). this is correct to use, right?
Thanks!
My function I'm using is: xtreg dependent_log_variable independent_binary_variable , re (also for fixed effects - xtreg dependent_log_variable independent_binary_variable , fe)
I was told that in this case I shouldn't get as result of the coefficient (beta) to be larger than 1, but I got 1.83. My data values seems to be ok.
My question is:
- is my beta coefficient of 1.83 sounds reasonable to get? if not - why shouldn't I get such result in this case?
- just to verify: in order to get the real value of the coefficient I'm using the follow equation: ( (exp(beta) - 1) *100). this is correct to use, right?
Thanks!
