Fatma Bircan wrote:
> Hello,
>
> I am doing multinomial logit estimates for the sector of choise using
> mothly data for Turkey. My multinomial logit estimates are sensible,
> the coefficients are of expected signs and mostly statistically
> significant. However, when I try to get the results for the marginal
> effects for each outcoume using "mfx compute, predict(outcome(1))"
> command, for example, I get only the marginal effects and means of the
> variables at which the marginal effects are evaluated and a note that
> "the second derivatives can not be computed". I also try to get the
> marginal effects using "dmlogit2" command. The marginal effect values
> using either command are the same for most of the variables and are
> different for some variables. Also, the t-ratios are 0 (zero) for all
> the marginal effects calculated using "dmlogit2" command. I do not get
> the t-ratios at all when I used "mfx compute" command. I do not expect
> 0 t-raios for the marginal effects.
You might be interested in my program margin. It computes marginal effects
analytically, thus you will definitely not met the error message.
To obtain it, visit the website http://www.bkae.hu/bartus/stata
or type in web-aware stata
net from http://www.bkae.hu/bartus/stata
net install margin
Hopes this help
Tamas
------------------------------------------------
Tamas Bartus, PhD
Assistant Professor, Department of Sociology and Social Policy
Budapest University of Economic Sciences and Public Administration
1093 Budapest, Fovam ter 8.
Phone: +36-1-217-51-72 Fax: +36-1-217-44-82
Homepage: http://www.bke.hu/bartus
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/