Yihui Xie


Simulation of QQ plots for the Normal distribution

Yihui Xie & Lijia Yu / 2017-04-04

This demo shows the possible QQ plots created by random numbers generated from a Normal distribution so that users can get a rough idea about how QQ plots really look like.

When the sample size is small, it is hard to get a correct inference about the distribution of data from a QQ plot. Even if the sample size is large, usually there are outliers far away from the straight line. Therefore, don’t overinterpret the QQ plots.

ani.options(interval = 0.1, nmax = 100)
par(mar = c(3, 3, 2, 0.5), mgp = c(1.5, 0.5, 0), tcl = -0.3)

sim.qqnorm(n = 20, last.plot = expression(abline(0, 1)))