R code for fixed effect multinomial logistic regression and experimental data
Hi R support group team
I would like to contribute on two issues posted last days:
1. Addressing question by Valerio [Valerio Leone Sciabolazza <
[hidden email]>]: [R] how to run a multinomial logistic regression
with fixed effects.
Application of fixed effect multinomial logistic regression (FE MNLR) has
many drawbacks as explained in literatures. This could be a reason that R
experts did not work on FE MNLR.
A. Green 2013. 15 Panel Data Models for Discrete Choice.
http://people.stern.nyu.edu/wgreene/Econometrics/Greene-PanelDataModelsforDiscreteChoice.pdf B. Xavier D’Haultfœuille and Alessandro Iaria 2015. A Convenient Method
for the Estimation of the Multinomial Logit Model with Fixed Effects.
2. Solutions my posts: a. Regarding my experimental data dealing problem in
R, I separately developed network graphs of treated and untreated
groups which made me easier to compare and understand the intervention
effect difference. b. Regarding my missing data handling problem, console
script accounted the missing data issue but I had forgoten filling "NA" in
a box of Rstudio while importing data.
Thanks for your attempt to resolve my R programing problem.