Canoco5 Support Site

Variation partitioning is easily accessible for two or three groups of predictors including calculations of individual fractions of explained variation, based either on partial or non-partial analyses and using either raw or adjusted variation estimates.

The above snapshot shows a specialized page shown in analysis notebook of a variation partitioning analysis. The variation in the response (community data) is explained by two groups of predictors and the partitioning procedure helps you to separate the unique contributions of these two groups (a, b) from their shared explanatory power. The individual fractions are described in the top table, while the central table contains tests of significance for the unique effects of the two groups and for their shared effect.

 

Copyright Petr Šmilauer © 2012-2021. All Rights Reserved.