I have an interaction plot that seems to be overlapping on top of one another, what does this signify?
Contents
hide
Best Answer
The straightforward interpretation is that at the high and low values of BWRatio the mean of "value" is higher for G than it is for N; whereas in the medium values G and N are virtually identical.
Whether this difference is either material or significant is another question which can't be judged from this plot.
The plot could be improved by adding the actual data to it, perhaps in light grey in the background. This would give more of an immediate idea of how tightly centred around the means the data are and hence a visual sense of how important the deviation of the two lines from eachother is.
Similar Posts:
- Solved – What are best practices in identifying interaction effects
- Solved – How to interpret interaction coefficients in a temporally lagged regression model
- Solved – Can interaction be statistically significant while only one of two main effect are significant
- Solved – How to deal with interaction term’s VIF score
- Solved – Does a factor-by-factor interaction term have any literal interpretation