1
$\begingroup$

I ran an experiment with 4 factors A, B, C, and D.

Factor C is nested within B.

The results of the ANOVA show:

C had a significant effect

and

AxB had a significant interaction

D had no significant effect, but in this case, D is species, and I want to compare these two species side by side.

So when I want to present these graphically, should I show:

a graph with C and D PLUS another graph with AxB and D

OR

A graph showing all four factors.

I ask because factor C has 9 levels, and I have 4 different response variables I'd like to show (length, height, width, mortality), so there are quite a few graphs that will be coming from this if I show all factors.

$\endgroup$
2
$\begingroup$

This may be a greater pain to explain and digest than to display, no reason why it cannot be a simple 2D bar plot e.g. by adding an extra factor to a layout similar to Fig. 3D in this example. If A and B had a significant interaction and C is nested within B, it would be more correct to show A, B, and C in the same figure, and either have a separate figure/panel for the two levels of D or have all 4 factors together. Alternatively, it can be represented using 3D tools (bars or heatmaps) or a combination of 2D bars (to properly show variability and interactions) and a heatmap to better highlight the strength of response patterns, which may be needed if the total number of factor combinations is very large. As a side note, morphometric measures tend to be highly correlated.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.