This may help:
http://stats.stackexchange.com/questions/6557/how-can-one-plot-continuous-by-continuous-interactions-in-ggplot2
Thanks Matthias. I have found some similar R plots before but I find the graphics a bit dissapointing (especially with publication in mind), for now I'm sticking with a split-up linear regression and devising excel plots according to the regression formula.
You'll want to look at the visreg package for easy plots of this type.
However, it's relatively trivial to generate the data to plot by making a new data.frame containing predictors of the appropriate kind. You will need one variable for the x-axis and then the other variable set to a constant. Using the predict function with your OLS model and that data.frame will generate the y-values for the lines. If making, say, 3 lines, I prefer to divide the data into thirds and colour the data to match the lines at the median of that section of the data. That way they all go on the same plot instead of multiple plots.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R Statistical Package