Page 1 of 1

Semi log model

Posted: Sat Dec 13, 2008 3:29 pm
by LM22
I need to estimate a semi log model. It looks like this: Ln(Wage)=β0+β1*Educ+β2*Exper+β3*Exper^2

Does anyone know what I should do with Ln in the estimation? Some of my co students think I should tape it in just like it is, but change Ln with Log. Will that be correct, or should I just tape it in like this: Wage=β0+β1*Educ+β2*Exper+β3*Exper^2, and use e on the numbers I get in the estimation?

The numbers after β are footnotes.

Re: Semi log model

Posted: Sat Dec 13, 2008 4:15 pm
by startz
I need to estimate a semi log model. It looks like this: Ln(Wage)=β0+β1*Educ+β2*Exper+β3*Exper^2

Does anyone know what I should do with Ln in the estimation? Some of my co students think I should tape it in just like it is, but change Ln with Log. Will that be correct, or should I just tape it in like this: Wage=β0+β1*Educ+β2*Exper+β3*Exper^2, and use e on the numbers I get in the estimation?

The numbers after β are footnotes.
Both ways are used. The first is more common, and the natural log function in EViews is log().