Ordination is a statistical technique that allows us to graphically summarize complex relationships. Also considered “dimension reduction,” these techniques offer a way to simplify many correlated variables (“dimensions”) into two or three, axes, which are must easier to interpret. In two dimensions we can then map our data points on a figure, and this often gives us greater insight into the data.
Ordination helps to separate strong patterns from weak ones, select the most important factors, uncover possible clusters, and suggests directions for further, more formal analysis. Ordination methods are often categorized into “indirect gradient analyses,” concerning only one group of variables, and “direct gradient analyses,” which consider the relationship between two groups of variables.
This workshop will introduce and compare common ordination methods (MDS, PCA, PCoA, CA) with real-world data examples and graphics. We will also discuss some of the common distance measures and when to use them.