(This article was originally published at Psychological Statistics, and syndicated at StatsBlogs.)
This is a quick follow up to my earlier post that discussed how to graph CIs for within-subjects (repeated measures) ANOVA designs. My forthcoming book Serious stats describes how to do this for between-subjects designs (a much simpler problem). The blog that accompanies the book now has a post summarizing the main options and explaining how to plot difference-adjusted CIs (95% CIs constructed so that non-overlapping intervals correspond to a statistically significant difference between means at p < .05). In addition, the post includes R functions to calculate and plot difference-adjusted CIs (though the calculations are not difficult to reproduce by hand).
UPDATE: I've now added functions for two-tiered CIs for between-subjects designs on the book blog. More generally my functions for the book, CIs for ANOVA and a few other things are all available here. I plan to update these functions regularly to add functionality and deal with any undocumented features.
Please comment on the article here: Psychological Statistics
