Fundamentals of Statistics contains material of various lectures and courses of H. Lohninger on statistics, data analysis and chemometrics......click here for more.


QQ Plot

Historically, QQ plots (quantile-quantile plots) can be traced back to normal propability plots. However, while normal probability plots allow only for normal distributions, QQ plots can be applied to arbitrary distributions.

In order to create a QQ plot the quantiles of the observations of the two samples to be compared are plotted against each other. If the observations belong to the same distribution, the data points should be distributed along a straight line.

Which quantiles to use for the creation of QQ plots largely depends on the number of observations. If the two samples have the same number of observations it is sufficient to sort both samples and plot the measured values against each other. If the sample sizes are not the same, the data points of the smaller sample can be used as "reference" quantiles. In this case the quantiles of the larger sample have to interpolated to the quantiles of the smaller sample.