st: Marginal effects of explanatory variables in ordered probit model using gllamm

[Wei Zhang <wzhang616@gmail.com>]

Dear Stata users,

I need some help on obtaining marginal effects of explanatory
variables. I have used "gllamm" to estimate an ordered probit model
for panel data. My dependent variable is the level of pest infestation
level ranking from 1 to 5:

gllamm y x1 x2 x3 x4 x5 x6 ..... , link(oprobit) i(id)

I am now trying to get marginal effects for the explanatory variables
similar to those reported in mfx/margins. I have tried "gllapred me,
mu marg above(1)", but I am not sure if it has worked. According to
the definition of the option "above", a single number can be given for
all ordinal responses. What does it mean? If I am to calculate the
marginal effects manually, should I use "gllapred me, xb" or "gllapred
me, u" to get "xbeta"? The definition in "help" isn't clear to me.

Thank you very much. I would really appreciate it if you could please
provide some help.

Wei Zhang
IFPRI
*
*   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/