Dear Statalisters,
I have a dataset consisting of the number of bat surveyed along 3km
transects. Each survey is repeated twice each year, which results in two
levels of clustering; surveys are nested within locations, and locations
are nested within years.
To model this data I am using xtmepoisson. However, on just over a third
of surveys no bat were observed. I believe this high level of zero counts
is causing the model to become unstable, as when I try and include more
than three explanatory variables I invariably get the error message
'Hessian has become unstable or asymmetric'
My question is this: Is there a way to model count data with a high
proportion of zero counts within the hierarchical modelling/xtme
framework, in much the same way that a zero inflated Poisson model would
do for non-clustered data?
Many thanks,
Katherine Boughey
--
Centre for Ecology, Evolution and Conservation
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/