Hello,
I'm trying to run a mlogit regression on my data, and have been unsuccessful
so far. The data I am working with consist of many observations of how
people react when given a certain number. I have just 2 data points per
observation: a number (there are ~300 different possible numbers) and then a
reaction (either 0, 1, or 2). The reactions are mutually exclusive and
exhaustive.
I am trying to devise models that can estimate the probability of the
reactions given the number. However, I don't seem to be able to run the
mlogit command on my data (mlogit.data() works fine). I've read through much
of the mlogit tutorial pdf and the examples, and I've done plenty of
searching here, but I haven't come up with a solution yet.
My question:
What format do I need to have my data in? Right now I have:
Number Reaction Individual Choice
Where each observation is repeated 3 times (For Reaction=0,1,2;
Choice=yes,no,no) and Individual and Number held constant in each set of 3.
So basically, I have a list 3x as long as my initial data list, that holds
values for all possible outcomes and whether or not they occurred.
I'm hesitant to ask additional questions until I've played around with
mlogit once my data is in the proper format, but if anyone has suggestions
for the mFormula that I should be using, that would be much appreciated.
--
View this message in context:
http://r.789695.n4.nabble.com/Single-Variable-mlogit-formatting-tp4121139p4121139.html
Sent from the R help mailing list archive at Nabble.com.
