The histogram of a set of numerical data (e.g. a set of marks) can have any shape. However some characteristics of the distribution are particularly important.

Centre
What is a 'typical value' near the centre of the distribution?
Spread
Are all values near this central value, or are they widely spread around it?
Skewness
Is the distribution of values fairly symmetric around the centre? Or do the values lower than the centre tend to be closer together than the values above the centre?
Clusters
Does the histogram have two or more distince peaks (called modes)?

In this chapter, we concentrate on the first two of these characteristics -- the centre and spread of the distribution. These are particularly important when two or more data sets are being compared.

Both centre and spread can be summarised with a numerical value — a summary statistic.