Plotting discrete variables?

[christopherhammerly]

Hi, I'm really happy with the package, and super excited about using it, but I'm having trouble getting an appealing visualization with my raw likert scale DV. I've attached the figure that I generated and the code. My main questions is whether there is a way to replace the bean plot with something like circles that grow bigger as the density of points on a given value increases? This would serve a similar function as the bean plot with a continuous DV.

Thanks again!

pirateplot(Response~Attractor+Grammaticality+Modifier,
data=data.acceptability,
jitter.val = .15,
inf.method = "se",
inf.within = Subject,
bar.f.o = 0,
inf.f.o = .5,
yaxt = "n",
gl.lty = 0,
cex.lab = .7)
par(cex=.7)
axis(side = 2, at = c(1,2,3,4,5,6,7),pos = -5.6, tick = F)
axis(side = 2, at = c(1,2,3,4,5,6,7),pos = .4, tick = F)

RDI-raw.pdf

[ndphillips]

Thanks for the note! I totally agree that the pirateplot does not look very nice with discrete DVs. Up until now, my answer has been that the pirateplot should only be used for continuous DVs, however, I can also see the benefit to adding an option for a discrete DV. And I agree that using circles is a nice idea. It would take some time for me to update some of the other elements, like the confidence interval and densities as these would now be based on something other than a normal distribution....l'll see what I can do. If you don't hear back from me in 2 weeks, feel free to nudge me!