Page 1 of 1

Interpreting results with a dummy at left hand side

Posted: Tue Aug 05, 2014 12:03 am
by ADNL
Hi,

In a model a regressed I used a dummy variable at the left-hand side to explain its presence. The explanatory variables at the right hand side are partly significant, but how do I interpret the coefficients.

For example: Y= b + 24.5 population + e

So one of the coefficients has a result of 0.245 for population. Is it possible to interpret it as: An increase of one unit of population causes an increase of 24.5% of the dummy variable. Or is it not possible and should it be interpreted differently. Or dummy variable is explained by population for 24.5% ?


Thank you in advance!

Re: Interpreting results with a dummy at left hand side

Posted: Tue Aug 05, 2014 6:35 am
by startz
Google "linear probability model"

Re: Interpreting results with a dummy at left hand side

Posted: Wed Sep 03, 2014 6:44 am
by Carlo Lazzaro
I'm not clear with your approach if your regresion setting is the linear one.
Conversely, if you are talking about a logistic regression a yes/no dependent variable makes sense. However, the interpretation of your results could be easier if you convert the dependent variable into an odds ratio.

Kind regards,
Carlo

Re: Interpreting results with a dummy at left hand side

Posted: Mon Dec 08, 2014 9:32 am
by ADNL
Thanks everyone for your help!