Dear all,

I have a cross sectional data with a variable showing number of siblings "n_siblings" for each individual. I also have a variable "m_fam" which identifies each family by a unique id number (so some individuals in the data are from the same family). I want to do a family and sibling fixed effects analysis. Please how do i go about this?

Will it be
xtset m_fam
clogit x y, group(m_fam)

or

xtset n_siblings
clogit x y, group(n_siblings)