Dear statlister,
I have a simple question regarding item-specific thresholds models
using -gllamm-. Why -gllamm- with -lv- option gives different
coefficients and different significance levels than -gllamm- with
-thresh- option (and -gologit2-)? And how should I interpret the
results in the two cases? According to the gllamm manual they are
supposed to be two ways of specifying the same model.
Mirko asked me this before privately and I had no idea, but I think
it just hit me. With the lv option, it is showing you what the
cutpoints are with each possible reference category. Ergo, the
_cut1x parameters are the cutpoints when category 1 is the reference
category (note how each of those corresponds to a constant term in
the first gllamm run). Then _cut2x are the cutpoints using category
2 as the reference (take the constants from the original gllamm run
and add the coefficient for d2). etc. You get 6 sets of cutpoints
because there are 6 possible reference categories. I assume this is
meant to be some sort of convenience, although my skimming of the
documentation didn't make it very clear to me either.