I am trying to run an RDA or CCA (Redundancy Analysis, And Canonical Correlation) to illustrate the relationship between measured gene abundance and a set of env. variables. The environmental variables are measured in greatly different scales (pH, oxygen concentration, Nitrogen concentration, Temperature....). I realized that I need to standarise my data after getting some strange results in the RDA. I was wondering if I could use excel to calculate a z-score (for each variable) by just grabbing all the data for each variable and follwing the normal procedure. Then I would use those z-scores for RDA directly. Could I do that for all env. variables at once? As I understand it, I dont need to standarize my gene abundance data (response variables) since its all on the same scale.
Yes, because z score standardise data to their means :)
The most effective way to standardize data from environmental variables expressed in different scales is to rescale the data so that it is expressed in the same scale.
You can read this article. It is very useful to select the appropriate multivariable method, and the "mathematical treatment" to employ your data.
Article Ramette A.. Multivariate analyses in microbial ecology. FEMS...
- Badges
- Science topic
- Similar topics
- Biological Science
- Microbiology