Are You Too A Community Ecologist? Use Primer? Then Go No Further!
For your information, there might be a link to a pdf of the copyrighted manual to widely used community ecology analysis program (click on picture below to download pdf). It might only be up for a short while, so you should probably download it if you use or plan on using Primer, or interested in community ecology analysis, relatively soon. It also might be around 17mb if such a document were to exist on the internet in violation of copyright. Of course I would never post anything on the internet that would break copyright law, open access to information is evil and undermines peer review yadda yadda....
Ha ha ha.
ReplyDeleteUniversity of Auckland huh?
Tut tut tut...
I bet I actually held the hard copy of this book in my hands two years ago.
It's like seeing an old friend again.
Actually, I was amused that the enormous Paterno Library at Penn State didn't have this ubiquitous text. But then we don't really have many community ecologists, at least in Biology. But of all the places the Interlibrary Loan could get the book from, It was the Kiwis that came through! They had to go across the entire world for it. Its not that rare is it??
ReplyDeleteThe only other copies I have seen have been ring-bound...
ReplyDeleteThey don't have a copy at the usually useful University of Melbourne library either.
I reckon the worst thing about PRIMER though is that most people never read the text - just use the defaults, click some buttons, and away you go. How many methods have you read where n=~100's and "a log transform was applied to the data before analysis"?
I couldn't imagine using a program where all the tests are weird acronyms. I found out I needed the manual when my advisor asked me what ANOSIM did after printed out nice looking analyses to show him. I just sat there and was like " um.. compares two things and sees if its similar?" I felt a little dumb needless to say, that I hadn't bothered to look at exactly how the test worked and what the output (aside from p-values and similarity percentages) actually meant.
ReplyDeleteWith that all being said, I can't imagine people publishing analyses of their data using tests they don't know the backbone of.
I also would recommend Biometry (no I will not scan it in lol) for a good lesson on the appropriate use of transformations.
p.s.-the log transform didn't work for my abundance data, but the quarter-root did...
re: transforming your data -
ReplyDeleteMore important when you are dealing with lots of micro-critters I guess, but not so much when it's only the big stuff you're counting.
And Sokal - I think of that book as "The Bible". But with less voodoo, obviously ;)
I am about to use PRIMER for first time in my undergraduare dissertation, so this should be really useful...thanks!
ReplyDeleteI am looking at the differences in fish assemblages and estimating population densities, between angled and non-angled sites in a marine reserve. Anybody know which stats would be best? I am reading different things everywhere!
Hi nat,
ReplyDeleteThe Primer manual is a great resource, but make sure you read the primary literature so you understand it in better detail.
I'm not sure of your details, but if you have replicates of multispecies communities with abundance data (quantitatively sampled), a bray-curtis MDS and cluster analysis would represent the similarity of the communities well. you can color code by factors (angled vs. nonangled) and bootstrap it up for some confidence intervals.