Q'Q Plots

Q'Q Plots (quantile-quantile) plots are found in the Graphs menu: .

This kind of probability plot plots the quantiles of a variable's distribution against the quantiles of a test distribution. Probability plots are generally used to determine whether the distribution of a variable matches a given distribution. If the selected variable matches the test distribution, the points cluster around the straight line.

When requesting a Q-Q plot, a second plot (not shown here) is produced with a detrended form, detrended meaning that you are concentrating on deviations from the normal (reference) distribution, instead of looking at the overall picture. The straight line in the plot represents the perfectly normal distribution.

By default SPSS uses the normal distribution.

The commands offers a number of advanced possibilities, namely checking various distributions, performing some transformation before doing so and options that affect the computation of fractiles.

There is also a P'P plot (), i.e. a similar idea, except instead of plotting quantiles against quantiles, it plots cumulative proportions of an empirical variable against cumulative proportions of a theoretical distribution.

You can also obtain normal probability (Q'Q) plots from the menu. You might want to use this command when requesting Q'Q for groups, e.g. continents (factor variable).