Travis, I would suggest reading over section 18.1.3 in Quinn, G. P and M. J. Keough. Experimental Design and Data Analysis for Biologists. Cambridge University Press, Cambridge, UK. It has a nice discussion of the various methodologies available for testing hypotheses about group differences based upon similarities as opposed to variances. It includes discussions of ANOSIM, MRPP and distance based redundancy analysis. If you have one-way models to test than any of the aforementioned procedures can be used. However ANOSIM and MRPP are limited to one-way models while redundancy analysis can be used for any model for which a design matrix can be specified.
As for any statistical test, redundancy analysis will provide a p value (permutational or Monte Carlo based) that can be used to assess your confidence that a difference exists between the parameters of the populations based upon the statistics generated from the samples (i.e. statistics, if done properly, allow us to enlarge our inference space beyond our samples). I strongly disagree with many of the statements expostulated on this forum on the uselessness of p values. It represents a fundamental misunderstanding of why we perform statistical tests and a misconception that statistical analysis is limited to Popperian hypothesis tests, which could not be further from the truth. Jim ><(((º> ><(((º> ><(((º> ><(((º> ><(((º> ><(((º> Jim Novak Biological Sciences Department 1162 life Sciences Annex Eastern Illinois University Charleston, IL 61920 (217) 581-6385 (217) 581-7141 (FAX) [email protected] http://www.ux1.eiu.edu/~jmnovak/ ><(((º> ><(((º> <º)))>< ><(((º> ><(((º> ><(((º> Confidentiality Notice: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message. On Mar 30, 2011, at 8:08 AM, Palmer, Mike wrote: > Travis, > Note that partial Canonical Correspondence Analysis and partial Redundancy > Analysis are ideal for this sort of a hypothesis test (especially if the > permutations are set up correctly). Also note that it is not 'ordination vs. > MRPP', because these direct gradient analysis techniques, when applied to > categories, can be considered special cases of MRPP: Palmer, M.W., D.J. > McGlinn, L. Westerberg, and P. Milberg. 2008. Indices for detecting changes > in species composition: some simplifications. Ecology. 89:1769-1771. > Direct gradient analysis such as pCCA or pRDA yields diagnostics for species > and samples, and allows graphical representation of results. You get far > more than just a p-value. While the job of interpreting results is always up > to the interpreter, these diagnostics can help get you on a solid footing. I > suggest avoiding NMDS, as this technique collapses information about species > identity - and not being a direct gradient analysis technique, it is of > questionable value when you have clean hypothesis tests. > ---Mike Palmer > ________________________________________ > From: Ecological Society of America: grants, jobs, news > [[email protected]] On Behalf Of Liz Pryde [[email protected]] > Sent: Wednesday, March 30, 2011 4:01 AM > To: [email protected] > Subject: Re: [ECOLOG-L] Community Analysis Hypothesis Tests > > Hi Travis, > > As it is a community comparison I think ANOSIM is still widely accepted to > test for differences between treatments/time, particularly in the marine > literature. Manuel is correct, it will not tell you the biological reason > behind a difference and cannot really give a magnitude of difference per se > (depending on the design of the study) but it does give a meaningful p value > based on permutations of configurations of the presence-absence matrix. In > this way its limitations are analogous with ANOVAs. As this is a comparative > study I would think that a statistically significant p value would be > meaningful. > > However, teasing out the question "why the difference" requires closer > examination and comparisons between the community matrices and say, > environmental variables. Examples of these types of further analyses can be > found in the PRIMER manuals (ver 6) and in the extensive literature that can > be found on the PRIMER-E website. > > Hope that helps, > Liz > > > > On Tue, Mar 29, 2011 at 11:12 PM, Manuel Spínola <[email protected]>wrote: > >> Hi Travis, >> >> I don't think that a p-value is going to tell you if there is a biological >> meaningful difference between community. >> >> What will be the metrics that you are planning to use with ANOSIM ? >> >> Without seeing the data I can tell you that there is a difference between >> communities, but the important question is how different they are, so you >> can assess a practical or biological significance and a p-value is not going >> to tell you that. >> >> Best, >> >> Manuel >> >> >> On 28/03/2011 07:14 a.m., T. Travis Brown wrote: >> >>> Hello, I am trying to determine the best way to test for a difference in >>> the overall mussel community found in a stream between 1980 and 2008. I >>> have seven sites with presence/absence data. In addition to various >>> descriptive statistics and graphs (nonmetric multidimensional scaling) I >>> would like to use ANOSIM because it offers a P-value, and answers the >>> question: "well, is there a difference or not?". I am not as up-to-date on >>> this literature as I would like to be. Does anyone know if this is still an >>> accepted test? Would some type of multi-response permutation procedure be >>> better? >>> >>> >>> T. Travis Brown >>> [email protected] >>> >>> >> >> -- >> *Manuel Spínola, Ph.D.* >> Instituto Internacional en Conservación y Manejo de Vida Silvestre >> Universidad Nacional >> Apartado 1350-3000 >> Heredia >> COSTA RICA >> [email protected] >> [email protected] >> Teléfono: (506) 2277-3598 >> Fax: (506) 2237-7036 >> Personal website: Lobito de río < >> https://sites.google.com/site/lobitoderio/> >> Institutional website: ICOMVIS <http://www.icomvis.una.ac.cr/> >> > > > > -- > Liz Pryde > PhD Candidate (off-campus) > School of Earth and Environmental Sciences > James Cook University > > Thornbury, Melbourne > 0418551570
