Thanks to Kit Baum a new package is available on SSC: -fmlogit-.
-fmlogit- fits by quasi maximum likelihood a fractional multinomial
logit model. It models a set of dependent variables that each must
range between 0 and 1 and must always, for each observation, add up to
1: for example, they may be proportions. It is a multivariate
generalization of the fractional logit model proposed by Papke and
Wooldridge (1996).
-fmlogit- is an alternative to -dirifit- by me, Nick Cox and Stephen
Jenkins and is also downloadable from SSC. -fmlogit- may be
particularly useful in large dataset when some of the proportions that
are being model are either zero or one, and there is nothing special
about those zeros and ones, for instance they occurred through rounding
during measurement. Sometimes, the zeros and ones occur through a
special process, for example the proportion time spent taking care of
children will have a reasonably smooth distribution well away from zero
except for a spike at zero for those persons without children.
-fmlogit- cannot take care of those situations.
To install -fmlogit- type -ssc install fmlogit-.
-- Maarten
Papke,Leslie E. and Jeffrey M. Wooldridge (1996), Econometric Methods
for Fractional Response Variables with an application to 401(k) Plan
Participation Rates. Journal of Applied Econometrics 11(6):619--632.
PS. those who are worried about my dissertation progress: I just
finished two chapters, so I decided that I could indulge in an
afternoon of programming.
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room N515
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
*
* 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/
