How to communicate stats

#1
Hi. I'm new and hope someone can help me. I work in soil microbiology field. I have my stats done, (know more about the inside of my car than of stats...) and then I get hundreds of pages to sift through before I do my report/article. My problem: How do I relate interaction in ANOVA in laymen's tearms? I do know it is when the different factors have a bigger influence than the sum of the total factors, but how to I explain it? The same with PCA - I know the different directions of the lines show positive (same direction) or negative (opposite direction) correlation with the longer lines the bigger correlation. But HOW do I explain that?
 
Last edited:

ledzep

Point Mass at Zero
#2
Interactions are generally nightmare for everyone (especially if there are >=3 way interactions).

Say you found the interaction between two factors, A and B significant. This means that the effect of factor A depends on the level of factor B.
For example: http://wilderdom.com/statistics/Interaction.html

If you look at graph (e) in the above example, the interaction between sleep and intensity of exercise is significant (because the two lines crosses). This means that the effect of sleep depends on the intensity of exersise (i.e. effect of one factor depends on the level of another factor). The effect of sleep is high at light level of exercise, ad low at heavy level of exercise.
 

Karabiner

TS Contributor
#4
An interaction is a difference between differences. Therefore, a 3-way interaction is a difference of the difference between differences.

Kind regards

K.