Can you try to roll back to 3.5?
Best,
Kosuke
Department of Politics
Princeton University
http://imai.princeton.edu
On Sep 5, 2013, at 10:11 AM, Eva Bernauer
<eva.bernauer@uni-mannheim.de<mailto:eva.bernauer@uni-mannheim.de>> wrote:
Dear Kosuke Imai,
may I ask you a question on Zelig? (My mail to Matt Owen somehow could not be delivered).
I found that the logit and the relogit model in zelig do not appropriately run
with robust standard errors. In fact, the results do not change when robust=TRUE. As you
might know, this issue has already been discussed in the Zelig forum.
Is there any solution to the problem?
I would be very glad, if you could tell me, as such a fix would faciliate everything so
much!
Best,
Eva