Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
Maarten Buis <maartenlbuis@gmail.com> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
st: Re: Question about ln-linear models |

Date |
Fri, 2 Sep 2011 15:31:48 +0200 |

On Fri, Sep 2, 2011 at 3:22 PM, Mihaylov, Emil wrote me privately: > I am STATA user and would like to aks you an econometric question. Thank you very much in advance for your time and help. I really apprecaite it. > My question is about ln-linear models, and more specifically how to read coefficients in these models. There are obviously two ways to interpter coefficients in ln-linear models and I am a little bit confused which one to use. Some econometric books suggest to read the coefficients in these models as (100*Beta)%. At the same time I have seen also many empirical works where coefficients are interpreted as exp(Beta)-1)*100. I suppose that both ways are correct, they just take a different perspective. Now I am wondering how to interpret the coefficients of my regression. I would apprecaite if you could help me. > > I have the following situation: lnY = lnX1 +…+ Dummy > > My dependent variable is the natural logarithm of the trade between country i and country j and the dummy variable indicates whether the two trading countries share the same origin of the legal system (same origin =1). I have estimated the model using OLS. The coefficient of the dummy variable is 0.85. So, my interpretation is that having the same origin of the legal system, all else being equal, would lead to 85% increase in trade. I am wondering however if this is correct. A friend of mine suggested that I have to read this as exp(0.85)-1)*100%. > > Could you tell me please which of both ways is correct in this situation? Thanks a lot in advance for your help. Such questions should not be sent to persons privately but to the statalist as is clearly explained in the Statalist FAQ: <http://www.stata.com/support/faqs/res/statalist.html#private> The two types of interpretation you discussed are not the same, they are appropriate for different models, see e.g.: http://blog.stata.com/2011/08/22/use-poisson-rather-than-regress-tell-a-friend/ -- Maarten -------------------------- Maarten L. Buis Institut fuer Soziologie Universitaet Tuebingen Wilhelmstrasse 36 72074 Tuebingen Germany http://www.maartenbuis.nl -------------------------- * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**Follow-Ups**:**Re: st: Re: Question about ln-linear models***From:*Nick Cox <njcoxstata@gmail.com>

- Prev by Date:
**Re: AW: st: nearmrg for strings (titles)** - Next by Date:
**Re: st: Package -ghansen- now available in SSC** - Previous by thread:
**Re: AW: st: nearmrg for strings (titles)** - Next by thread:
**Re: st: Re: Question about ln-linear models** - Index(es):