Taming the Analysis of Variance

Data vary. For example, rice yields differ from one field to another. ANOVA helps you investigate whether some of this variation is related to the characteristics you have measured.

A model is a simple description of the pattern of the data. The simplest model might use one predicted yield for every field. A more detailed model might allow yields to differ by variety, village, fertiliser use or field size.

ANOVA helps us compare these possibilities. It separates the variation explained by the model from the variation left unexplained. It asks whether adding a feature explains enough to improve the model. This helps you find a model that explains the data well without being more complicated than necessary.

Check out the document below;

Taming the ANOVA