Resource issues	Statistical solutions
“The issue was always resulting from the balance between what is	“They [the reviewers] were too focused on the possibility and
reasonably possible and what is ideal… People scream	effects of pseudoreplication than our approach to dealing with it”
pseudoreplication when it's not pseudoreplication”
	“Hurlburt did us all a disservice when he said that statistics can't
“…pseudoreplication is an issue that's been blown way out of	be used when pseudoreplication is present. They can but what
proportion. The real issue is how you interpret your results and	they tell you is something that warrants careful interpretation.”
then report them. How do you replicate things like marshes, forest
patches? You have to say, what I found is true for this forest and	“Pseudoreplication is a bogus term for a poorly nested design. The
then the question is how representative that forest is of all the forests in the area”	Hurlbert publication is one of the most pernicious publications in all of ecology.”
“Ultimately all field work is pseudoreplicated, depending upon	“Pseudoreplication is just a question of correct model
scale. I have been criticized because all my work occurred in 1	specification. If the model correctly reflects the sampling design,
estuary, 1 only in the Gulf of Mexico”	then the issue becomes one of parameter estimation and
	potential parameter nonidentifiability.”
“Often there is simply not the funding to conduct landscape scale
experiments without pseudoreplication”	“I would say “generalized linear multilevel models” but, yes,
	basically random effects”
“Replication is not always possible in ecology. This is particularly
true in restoration ecology when restored ecosystems are created	“I used multiple control sites, to at least differentiate the
at great expense and cannot always be replicated for the purpose	treatment area from multiple other sites.”
of scientific study. Sometimes we just have to study what is
there!”	“Indicating that samples were taken at a distance greater than the
	autocorrelation distance for many soil variables (from the
“In many ways, it is very difficult to really meet the needs of	literature in the same ecosystem) and framing the results and
replication and even when we do it is often somewhat arbitrary. In	conclusions to this experiment design.”
many instances, the research we have done could be better
referred to as ‘case studies,’ but then we'd have to pray a journal
will accept that.”
“Ecologists are so hung up on pseudoreplication, it's not even an
issue for my hydrologist colleagues, in whose research
pseudoreplication is often inevitable.”