[R] Question about food sampling analysis

[kMan]

Dear Sarah,

[snip...]
"I know that samples within each facility cannot be treated as independent,
so I need an approach that accounts for (1) clustering within facilities
and"

You could just use lm() & some planning. The data from within a specific
facility can be fit with a model to generate parameters that are compared
between facilities. Not to practical though - assuming the 57 production
facilities each have their own analytical lab, you'll have 57 different fits
to get your parameters from to use in your between test. Questions about
dependent data are fairly common, so it should be relatively straight
forward to get a solution and/or idea for a suitable package from the
archives.

"(2) the different number of samples taken at each facility."
It's a waste of time to worry about that. You'll be comparing aggregate
values between groups, and you'll have too few data-points within a group to
detect within effects... 

[snip...]

Sincerely,
KeithC.