st: seqlogit: Tree Specification

[Dominik Becker <dominik.becker@wiso.uni-koeln.de>]

Dear list (and, especially, dear Maarten Buis),

I would like to use the seqlogit package to model educational
transitions. The example given in the technical description of the
package (http://www.maartenbuis.nl/dissertation/tm_II.pdf)
assumes that in the first transition students can choose between less
than high school (1) and all other levels; the second transition is a
choice between leaving after high school (2) versus going to college (3
and 4); and the third transition is a choice between some college (3)
and a four-year course.

In contrast to this example, the German educational system distinguishes
between a high school degree ("Abitur") as the prerequisite for academic
studies at a 'conventional' university, and a vocational diploma
("Fachabitur"), which allows academic studies at a university of applied
sciences ("Fachhochschule"). Thus, for the German case the transition
decisions would be (at least!) the following:

d1: leaving the educational system after 10th class (0) vs. all other
alternatives (1 2 3 4 5)
d2: for those who passed their vocational diploma ("Fachabitur"):
leaving it by this (1 2) vs. go for "Abitur" (3 4 5 )
d3: for those with only Fachabitur: leaving the educational system (1)
vs. starting studies at a university of applied science (2)
d4: for those with Abitur: leaving the educational system (3) vs.
starting studies at a university of applied science (4) vs. staring
studies at a conventional university

My intention was to model the decision tree in the following way:

    tree (0 : 1 2 3 4 5 , 1 2 : 3 4 5 , 1 : 2 , 3 : 4 : 5)

However, Stata reports the error "No path in the tree leads to values 5.
r(198);"

The seqlogit help file says, "all levels must be reachable through one
and only one path through the tree". Maybe this condition is not
fulfilled for my model?

Thanks for all answers!

Best,
Dominik Becker


*
*   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/