st: 'ologit' and 'esttab'

[Mosi Ifatunji <ifatunji@gmail.com>]

Colleagues,

I am trying to generate a table of ordinal logit regressions ('ologit') using 'esttab.' The problem seems to be the bottom portion of the output associated with 'ologit' - namely, the cut points (or thresholds). This second portion of the output table seems to confuse 'esttab' resulting is a less than clearly formatted table (when the 'esttab' table is reported in the .smcl format).

My questions:

1. Is is possible to suppress reporting of the cut values from model estimation?

2. Is tere a way to incorporate the second portion of the table into the 'esttab' command so that the 'esttab' table is generated cleanly?

I have rerun the models using 'omodel logit' (which reports the cut values within the portion of the table were the coefficients for the variables are reported) and had no problem with table generation using 'esttab'. The problem is that I am generating my estimates using multiply imputed data from a complex survey. Unfortunately, 'mi estimate' does not particularly like the 'omodel logit' sequence. It would appear that I am stuck using 'ologit' (unless there is another way to generate ordinal logits using complex data (i.e., 'svy' friendly) across multiply imputed datasets (i.e., mi estimate friendly).

Any help would be greatly appreciated,

-- Mosi
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/