st: difference between interaction created with # and created with xi3

[uso stata <usostata@gmail.com>]

Hello statlist,
I am currently using a linear model specification and I am fiddling
with interactions.

I have three BINARY regressors a b c

When I use the command a#b#c stata spits out 7 interactions:

a b c
0 0 1
0 1 0
0 1 1
1 0 0
1 0 1
1 1 0
1 1 1

when I use xi3: and a*b*c I get:
a
b
c
_Iaxb
_Iaxc
_Ibxc
_axbxc

While the main effects are identical, i.e. the coefficient of "0 0 1"
is the same as the coefficient of "c", the interaction terms are all
different.

P

The xi3 methods generates the same coefficient as the a##b##c method,
that is equivalent to i.a i.b i.c a#b#c

So my question is what is the specification that stata creates for
a#b#c? How do I interpret for example "0 1 1", I thought it was the
same as the interaction of b and c, but it seems not.
*
*   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/