Checking for IIA violation in R

[mimii.zai]

Hi all,
I am such a beginner in this and I'm super stressed out about my data analysis and could use some help.
So I wanted to check if the IIA assumption holds for my Conditional logit model in R but I cannot code to save my life, which makes me wonder why I am even doing a choice experiment. But I am eager to learn!!
If anyone has experiences coding in R and could help me out with some lines of code then I would be forever grateful.

P.S I am using the GMNL package

Yours sincerely,
A confused coder

[Michiel Bliemer]

I have never checked IIA, I generally assume that IIA is violated and therefore use a mixed logit or nested logit model. Mixed logit models can account for the panel nature of the data (since a respondent is typically given multiple choice tasks in a choice experiment and therefore there exist correlations in the data within each respondent) and therefore is often the model of choice. A conditional logit model is typically only a starting point when modelling.

I am not familiar with the GMNL package, but you can consider using Apollo (also in R), which has quite an extensive manual and provides example syntax.
http://www.apollochoicemodelling.com/index.html

Apollo also has an online forum where you can get assistance.

Michiel