MerQuR+Statistics Glossary

Statistics Glossary → Map classification methods (equal interval, quantile, natural breaks)

Map classification methods (equal interval, quantile, natural breaks)concept

Rules that decide how continuous values on a thematic map are divided into colour classes.

The equal-interval method divides the value range into classes of equal width (for example 0-20, 20-40, …); it is easy to read, but with skewed data most observations end up in one or two classes. The quantile method puts roughly the same number of units in each class; colours are spread evenly, but very similar values can land in different classes and very different values in the same one. Natural breaks (Jenks) look for boundaries that minimize within-class variance and maximize between-class differences; Fisher's dynamic programming algorithm finds these optimal boundaries exactly.

The same data can give very different impressions under different classifications; whether a problem looks 'widespread' or 'local' can depend on the method chosen. The number of classes matters too: too few classes hide information, too many make colours hard to tell apart. Because Jenks boundaries are specific to the data, they are not suited to comparing maps from different dates; fixed boundaries are preferred for comparisons over time.

An everyday example

The population density of 50 villages in a province is mapped; most villages have 10-50 people per km², while two district centres exceed 800. With 5 equal-interval classes, 48 villages share the same pale colour. The quantile method puts 10 villages in each colour and shows the differences between villages, while Jenks places the two centres in a class of their own and splits the rest into natural groups.

In MerQur

The 'Classification (Continuous colour, Equal interval, Quantile, Natural breaks (Jenks)) and number of classes' option in the Hexbin Map; the legend shows the class boundaries and the number of cells in each class.

Analyses where this concept appears

Hexbin Map

Related concepts

Percentiles (quantiles) and percentile definitions