|Top Previous Next|
This menu allows you to:
(1) undertake an Analysis of Similarity (ANOSIM),
(2) identify those species which most contribute to the similarity (or dissimilarity) between samples (SIMPER), and
(3) undertake a Discriminant or Canonical Variates Analysis.
For video demonstrations see the Help: Guides: Grouping from Form.
Why does CAP allow you to assign samples to groups? You might for example allocate a series of samples to polluted and control groups, or in a benthic study, to different physical conditions such as mud, sand and gravel substrates. In another type of study you might want to define as a group all samples that held a species of particular interest.
By allocating samples to groups you can show the different groups in ordination plots and other graphs.
For example, in the hinkley fish.csv example dataset supplied with CAP, the species of fish caught in different years are grouped as pre- and post-1986. If a PCA is run on these data the two groups can be clearly seen to cluster around different positions within the ordination space. Note that the data were square-root transformed to get the clear result shown.
Once samples have been placed into groups you can test if the similarity of samples within each group is greater than the similarity that would occur by random chance, using ANOSIM, or you can undertake a Discriminant Analysis.