The shape of a numerical data distribution is mostly simply described as symmetric if it is roughly evenly spread around some central point or skewed, if it is not. If a distribution is skewed, it can be further described as positively skewed (‘tailing-off’ to the upper end of the distribution) or negatively skewed (‘tailing-off’ to the lower end of the distribution).

These three distribution shapes are illustrated in the parallel dot plot display below.

Dot plots, histograms and stem plots can all be used to investigate the shape of a data distribution.