Reading xtlogit re Code

[Helena Gruener]

Dear Stata-Users,

I found the following code for a xtlogit-model and (after reading the xtlogit manual) I'm failing to fully understand it:

xtlogit y x1 x2 x3, re i( countryid)


What does the underlined part of the code tell me (i(countryid))? Has it to do with clustering or does it relate to the re-model specification?

Thank you in advance and best regards,

Helena Gruener

[Carlo Lazzaro]

Helena:
-i(countryid) allows you to run a panel data regression even if your data were not previously -xtset-.
It works like:

xtset coutryid

As a consequence (and consistently with the code above), -i(countryid) does not include any -tempvar-: hence, time series operators are not allowed.

[Helena Gruener]

Thank you very much for your useful answer!