Jump to main content.


Model Report

Mean Similarity Analysis Version 6

Last Revision Date: 08/31/2009 View as PDF
General Information Back to Top
Model Abbreviated Name:

Model Extended Name:

Mean Similarity Analysis Version 6
Model Overview/Abstract:
MEANSIM6 contains software for Mean Similarity Analysis, a method of assessing the strength of a classification of many objects (sites) into a relatively small number of groups. Classification strength is measured by the extent to which sites within the same groups are more similar to each other, on average, than they are to sites in different groups.
Model Technical Contact Information:
John Van Sickle
US EPA, National Health and Environmental Effects Research Laboratory
Western Ecology Division
200 SW 35th St.
Corvallis, OR 97333
ph: 541-754-4314
fax: 541-754-4716
Model Homepage: http://www.epa.gov/wed/pages/models/dendro/mean_similarity_analysis.htm

User Information Back to Top
Technical Requirements
Computer Hardware
Desktop Windows PC
Compatible Operating Systems
Windows 2000, XP
Download Information
Download from http://www.epa.gov/wed/pages/models/dendro/mean_similarity_analysis.htm
Using the Model
Basic Model Inputs
Flat file containing a matrix of between-object similarities and the groups to which the objects belong.
Basic Model Outputs
Screen printout of mean within and between-group similarities, and results of a permutation test of their equivalence.
User Support
User's Guide Available?
Available at: http://www.epa.gov/wed/pages/models/dendro/mean_similarity_analysis.htm

User Qualifications
College degree, ability to calculate and interpret measures of similarity or dissimilarity.

Model Science Back to Top
Problem Identification
Species assemblages (fish, macroinvertebrates, periphyton, plants) are valuable indicators of the ecological health of aquatic ecosystems. Many quantitative measures are available to describe the taxonomic similarity or dissimilarity between 2 sampled assemblages. However, these similarity measures are not equivalent to Euclidian distance, which means that conventional parametric statistical methods cannot be used to analyze them. This program applies special methods to statistically test similarities. It calculates the mean similarities between and within pre-defined groups of multiple sampled assemblages, where groups might represent geographic regions or classes of site impairment. The program then tests whether assemblages within the groups have significantly smaller mean similarity than that of assemblages in different groups. The test is analogous to a 1-way MANOVA for the difference between assemblages.

Summary of Model Structure and Methods
Computes mean within and between group similarities and carries out a permutation/randomization test.
Model Evaluation
Van Sickle, J. 1997. Using mean similarity dendrograms to evaluate classifications. Journal of Agricultural, Biological and Environmental Statistics 2, 370-388.

Key Limitations to Model Scope
Upper limits of 5000 x 5000 similarity matrix and 50 groups.
Case Studies
Van Sickle, J. and R.M. Hughes. 2000. Classification strengths of ecoregions, catchments, and geographic clusters for aquatic vertebrates in Oregon. Journal of the North American Benthological Society 19, 370-384.

Local Navigation

Jump to main content.