To show the distribution of values in small data sets, graphical displays that separately represent each value are usually best. For example, dot plots and stem-and-leaf plots of marks from a single class identify the marks of individual students as well as showing the overall distribution.
However, for large data sets, we are usually less interested in individual marks than in getting a picture of the overall distribution. Indeed, separately representing thousands of individual values in a graphical display can make it harder to assess the shape of the distribution — there is too much 'noise' in the display that makes the 'signal' less prominent.
In this chapter, we consider graphical displays that summarise data. Some information is lost (individual values), but the important distributional information is shown more clearly.