Command for fixed effects ordered logit

[paulvonhippel]

Is there a user-developed Stata command for fixed effects ordered logit? I know that xtologit does not implement it. Code has been posted here for the BUC algorithm of Baetschmann et al., but I haven't seen a version that's been tested, documented, and distributed through SSC or another outlet.

[ericmelse]

Dear Paul,

I can recommend the following:
Stata UK 2011 A Dickerson AR Hole L Munford A review of estimators for the fixed effects ordered logit model
I assume that you are familiar with their paper:
A Dickerson AR Hole L Munford (2014). The relationship between well-being and commuting revisited. Does the choice of methodology matter? Regional Science and Urban Economics 49, 321-329, that is available here.

Two other papers might be of interest as well:

Muris, Chris (2017). Estimation in the Fixed Effects Ordered Logit Model. Review of Economics and Statistics 99:3, 465-477, institutional access here, with (Stata) replication materials available here. A presentation by Chris Muris (2016) is available here. This was discussed earlier on The Stata Forum here.

TL Liddell, JK Kruschke (2018). Analyzing ordinal data with metric models What could possibly go wrong? Journal of Experimental Social Psychology 79, 328-348, institutional access here, with a video talk available here, and (R) replication materials available here. Bruce Weaver discussed this earlier on The Stata Forum here.

All of the above is not exactly what you are looking for, but it is possibly of some help.

[paulvonhippel]

Thank you, Eric. At this point I know a bit about the fixed effects ordinal logit estimators that have been proposed in the literature. What I'm looking for is a documented Stata command that implements one of them. Cutting, pasting, and editing code from someone's replication package is not as useful as having a command that is documented, has been tested, and works for a variety of problems. We have tried code from a replication package and have found it does not return the expected results for our data.

An SSC command for FE ordinal logit would be a great service to the community. Does one exist? Is anyone working on one?

[Richard Williams]

You might consider the xthybid command. Allison's comments are at

https://statisticalhorizons.com/prob...-hybrid-method

I have some notes and references here:

https://www3.nd.edu/~rwilliam/Taiwan2018/Hybrid.pdf