I'm running an logit model using the Zelig package. I get the following error...what could be wrong?
anes96two <- zelig(trade962a ~ age962 + education962 + personal962 + economy962 + partisan962 + employment962 + union962 + home962 + market962 + race962 + income962, model="mlogit", data=data96)
#Error in attr(tt, "depFactors")$depFactorVar :
# $ operator is invalid for atomic vectors
Are you sure your dependent variable is as factor? I think the error is because your dependent variable is not a factor. Try str(data96), or summary(data96) to see if trade962a has levels and how many levels. If not, then, just set as a factor. For example: data96$trade962a.f = as.factor(data96$trade962a)
Finally, if you wanna run a logistic regression, you should use 'model="logit"', not "mlogit". Mlogit is for multinomial logit. Thus, if your variable is binary, using mlogit may cause the cited error. If so, just use model="logit".
If none of the above solve your problem, try to plot your dependent variable to see if the data is all right.
regards, Manoel