Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: gllamm missing cut points
From
Stas Kolenikov <[email protected]>
To
[email protected]
Subject
Re: st: gllamm missing cut points
Date
Sat, 9 Jun 2012 07:48:12 -0500
As a thought -- your rare category may be confounded with one of the
random effects. I.e., all of the rare categories were marked off by a
single subject, and that's the only category she selected. In this
case, this parameter is not identifiable, and -gllamm- may be just
smart enough to figure it out and drop it (or the whole group of
observations associated with that subject).
On Fri, Jun 8, 2012 at 10:02 AM, Joseph Padgett <[email protected]> wrote:
> Yeah. My first thought was that there must not be any respondents who
> checked off one of my categories. I ran some tabs using my estimation
> sample and there are responses in each category. Now, the lowest
> category has very few responses, but it seems like any responses at
> all should have made a difference. Also, what about gllamm dropping
> the first category? Does this mean that it dropped the second
> category because of too few responses in the first?
>
> I am in the process right now of going back through all of my coding
> checking for errors in each step because the nature of the error seems
> to me that may be something in my coding rather than the program's
> behavior as it seems to run fine other than the missing cut point.
>
> I code a lot of stuff (like my control variable sets, etc.) into
> macros so that I can write all of my models with the macros and then,
> should changes be necessary, I rewrite single macros instead of a
> dozen different programs. I'm going to weed my way back through those
> just to be sure everything is kosher. Hopefully I'll find whatever
> stray comma or cash symbol that is misbehaving and punish it dearly.
>
> I tossed this up on the off chance that this is a known issue because
> I feel like, with the male model running fine, that the coding must be
> in order since the variables are identical for each group
> (male/female).
>
> Who knows... if I find an error in my coding I'll be sure to respond
> to this thread and do a mea culpa.
>
> - Joseph
>
> On Fri, Jun 8, 2012 at 10:46 AM, Joerg Luedicke
> <[email protected]> wrote:
>> On Fri, Jun 8, 2012 at 9:44 AM, Joseph Padgett <[email protected]> wrote:
>>> There are responses for every category of my dependent variable
>>
>> Have you checked that for your actual estimation sample?
>>
>> J.
>> *
>> * 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/
>
> *
> * 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/
--
---- Stas Kolenikov
-- http://stas.kolenikov.name
---- Senior Survey Statistician, Abt SRBI
-- Opinions stated in this email are mine only, and do not reflect the
position of my employer
*
* 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/