st: why the level 2 and level 3 residuals are the same after fitting gllamm model

[jl591164@albany.edu]

Dear all,

I fitted three level(level 1-time, level 2-child, and level 3-county)
logistic regression nmodel using -gllamm-. Then I used -gllapred- to
obtain residuals at each of the three level to evalute at the normal
distrution assumption of the model.

The syntax that I used are showed below. The problem is that STATA gives
the same resiudles at the level 2(lev2) and level 3(lev3). The discriptive
statistics and histograms of lev2 and lev3 are exactly the same.

Is my syntax wrong? Can anyone tell me what is the problem? Many thanks.

Junqing

. gllapred lev3, ustd
(means and standard deviations will be stored in lev3m1 lev3s1 lev3m2 lev3s2)

. gllapred lev2, ustd
(means and standard deviations will be stored in lev2m1 lev2s1 lev2m2 lev2s2)

. gllapred lev1, pearson
(residuals will be stored in lev1)
log-likelihood:-7031.3976

*
*   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/