I needed to produce a few a interaction plots for my book in R and, while the interaction.plot() function is useful it has a couple of drawbacks. First, the default output isn't very pretty. Second, it works from the raw data, whereas I often need plots from cell means. For teaching purposes it is quite common to produce plots without raw data (for hypothetical data or from published examples).

My first attempts at the plots involved setting them up element by element. Just going over some examples I decided to turn the basic plot (for a 2 x 2 ANOVA) into a simple function. Nothing fancy, just a regular interaction plot in black and white that I think is prettier than the SPSS, Excel or R defaults. At some point I may have a go turning it into a general I x J ANOVA plot (or maybe even add CIs, but I'll probably do that from raw data if I ever get round to it).


plot.2by2 <- function(A1B1,A1B2, A2B1, A2B2, group.names, legend = TRUE, leg.loc=NULL, factor.labels=c('Factor A', 'Factor B'), swap = FALSE, ylab= NULL, main = NULL){
group.means <- c(A1B1, A2B1, A1B2, A2B2)
if(missing(ylab)) ylab <- expression(italic(DV))
if(swap==TRUE) {
group.names <- list(group.names[[2]], group.names[[1]]) ; group.means <- c(A1B1, A1B2, A2B1, A2B2); factor.labels <- c(factor.labels[2], factor.labels[1])
}
plot(group.means, pch=NA, ylim=c(min(group.means)*.95, max(group.means)*1.025), xlim=c(0.8,2.2), ylab=ylab, xaxt='n', xlab=factor.labels[1], main=main)
points(group.means[1:2], pch = 21)
points(group.means[3:4], pch = 19)
axis(side = 1, at = c(1:2), labels = group.names[[1]])
lines(group.means[1:2], lwd = .6, lty = 2)
lines(group.means[3:4], lwd = .6)
if(missing(leg.loc)) leg.loc <- c(1,max(group.means))
if(legend ==TRUE) legend(leg.loc[1], leg.loc[2],legend = group.names[[2]],  title = factor.labels[2], lty = c(3,1))
}


Call the function by entering the four cell means in conventional order: A1B1, A1B2 and so on where A1B1 is the mean of level 1 of factor A at level 1 of factor B. You also need a two item list containing text strings of the two level names of each factor. For instance:

lev.names <- list(c('A1', 'A2'), c('B1', 'B2'))
plot.2by2(5,15,10,20, lev.names)
You can swap the axes by adding the argument swap = TRUE:

plot.2by2(5,15,10,20, lev.names, swap = TRUE)



The default factor names are 'Factor A' and 'Factor B', but these are over-ridden in the call:

plot.2by2(5,15,10,20,lev.names, swap = TRUE, factor.labels= c('Factor 1','Factor 2'))


You can also change the y-axis label with ylab or add a main title with main.  The legend can be dropped (legend = FALSE) if you don't want one or need it to be located outside the plot. To move the legend just specify coordinates with an argument such as leg.loc = c(1,10). You can also edit the source code directly. Here is an example with title and meaningful labels:


group.names <- list(c('placebo','drug'), c('male', 'female'))


plot.2by2(10,10,15,20, group.names, factor.labels=c('Drug', 'Sex'), swap = FALSE)

As this just uses basic plotting functions in R you can also manipulate the plot in other ways: adding lines with segments(), adding text with text() changing graphical parameters with par() and so on. Depending on your platform it is also easy to extract the plot as a .pdf or .jpg file. On a mac I save it as a .pdf file and open it in preview which allows me to save it as .png, .gif or whatever I need.
Posted by
Labels: ANOVA displaying means psychology R statistics
0

Add a comment

I have been thinking to write a paper about MANOVA (and in particular why it should be avoided) for some time, but never got round to it. However, I recently discovered an excellent article by Francis Huang that pretty much sums up most of what I'd cover. In this blog post I'll just run through the main issues and refer you to Francis' paper for a more in-depth critique or the section on MANOVA in Serious Stats (Baguley, 2012).
2

I wrote a brief introduction to logistic regression aimed at psychology students. You can take a look at the pdf here:  

A more comprehensive introduction in terms of the generalised linear model can be found in my book:

Baguley, T. (2012). Serious stats: a guide to advanced statistics for the behavioral sciences. Palgrave Macmillan.

I wrote a short blog (with R Code) on how to calculate corrected CIs for rho and tau using the Fisher z transformation.

I have written a short article on Type II versus Type III SS in ANOVA-like models on my Serious Stats blog:

https://seriousstats.wordpress.com/2020/05/13/type-ii-and-type-iii-sums-of-squares-what-should-i-choose/

I have just published a short blog on the Egon Pearson correction for the chi-square test. This includes links to an R function to run the corrected test (and also provides residual analyses for contingency tables).

The blog is here and the R function here.

Bayesian Data Analysis in the Social Sciences Curriculum

Supported by the ESRC’s Advanced Training Initiative

Venue:           Bowden Room Nottingham Conference Centre

Burton Street, Nottingham, NG1 4BU

Booking information online

Provisional schedule:

Organizers:

Thom Baguley   twitter: @seriousstats

Mark Andrews  twitter: @xmjandrews

The third and (possibly) final round of the workshops of our introductory workshops was overbooked in April, but we have managed to arrange some additional dates in June.

There are still places left on these. More details at: http://www.priorexposure.org.uk/

As with the last round we are planning a free R workshop before hand (reccomended if you need a refresher or have never used R before).

In my Serious Stats blog I have a new post on providing CIs for a difference between independent R square coefficients.

You can find the post there or go direct to the function hosted on RPubs. I have been experimenting with knitr  but can't yet get the html from R Markdown to work with my blogger or wordpress blogs.
1
Links
Links
Blog Archive
Blog Archive
Subscribe
Subscribe
About Me
About Me
Loading
Dynamic Views theme. Powered by Blogger. Report Abuse.