Dear Statalisters,
I am new to multi-level modelling, and would like some help disentangling
the clustered nature of my data.
I am modelling the relationship between number of bats counted and
temperature and region, using xtmepoisson.
I have 12 sites, which are sampled on two random occasions each year. Each
year every site is sampled, so year and site are crossed. As each site is
sampled twice each year, each site-year combination is represented twice
in my data set. What is the most appropriate way of including this in my
model? Do I need to alter my current syntax:
xtmepoisson bats temp region || _all : R.year || site:
For example, should I include an extra level of clustering, with occasion
clustered within site-year, or a random interaction term between site and
year?
Thank you,
Katherine
--
School of Environmental Sciences
University of East Anglia
Norwich
NR4 7TJ
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
