Mixed Effects Ordered Logit Model - Help / Literature

[Guest]

Hello,

I am analysing data on a multilevel scale (country, village, individuals) with an ordered categorial response variable. As I want to account for the fixed effects on the country- and village level I am using the meologit model. Unfortunately I could not find any literature in the style of Camerons "Microeconometrics Using Stata" so far.
Could you pleanse give me some hints on relevant literatue which at best also explains the Stata implementation?

Thank you!

[Maarten Buis]

First, meologit is a random effects model, not a fixed effects model.

Second, The Stata Press book for this type of models is:
Sophia Rabe-Hesketh and Anders Skrondal (2012) Multilevel and Longitudinal Modeling Using Stata, Third Edition. College Station, TX: Stata Press. http://www.stata-press.com/books/multilevel-longitudinal-modeling-stata/

[Guest]

Hello Maarten,

thanks for your reply. I thought that meologit is a mixed effects model which involves both random- and fixed effects.
Thanks for the book hint; it sounds very convenient!

[Maarten Buis]

Terminology for this type of model is unfortunately rather "mixed". In the terminology typically used by economists, which I take you to be given the first reference you reported, this is not a fixed effects model, but just a random effects model with the possiblitly to add multiple levels and make some of the effects random. However, it does allow you to estimate the effects of explanatory variables. Confusingly, in some other disciplines these effects are sometimes refered to as fixed effects, but they have nothing to do with the fixed effects economists talk about.

[Guest]

Ha, so it's again a case of terminology differences in different disciplines (yes I have a background in economics). Know of this regarding statistics in economics & psychology. Thanks for enlightening me.

So I am starting to wonder, if there is a fixed-effects ordered logistic regression model in a multilevel framework. Unfortunately I haven’t found anything like that in the Stata documents.

[daniel klein]

You might find this thread interesting.

Best
Daniel

[Phil Schumm]

So I am starting to wonder, if there is a fixed-effects ordered logistic regression model in a multilevel framework.

See the following, which includes Stata code:

Baetschmann, G., Staub, K. E., & Winkelmann, R. (2011). Consistent Estimation of the Fixed Effects Ordered Logit Model. IZA Discussion Paper No. 5443.

[Guest]

Thank you guys. You helped me with my research quite a lot!

[Ujjwal]

Dear Maarten, Thank you for your reply. Would you please tell me how we can calculate marginal effect after running this meoprobit?

[Carlo Lazzaro]

Ujjwal:
-help meoprobit postestimation- and related entry in Stata .pdf manual it's probably what you're looking for.