A graph is a diagram for presenting numerical data and showing the relationship between them.
Graphs are the best way to display data when it is more important to convey overall patterns in the data than the individual data values. Well-designed graphs exploit human pattern recognition skills to enable readers to quickly and easily understand often complex data relationships. For example, the height of side-by-side bars is automatically understood as comparisons in magnitude across groups.
- show values, proportions, relationships between and across data, differences, distributions, ranges, aggregates, ranking, outliers and exceptions, deviations, correlations, confidence, and filters and highlights
- have a wide range of visual forms suited to different types of data and relationships
- make large datasets coherent
- are usually the most effective way of showing trends, differences, exceptions and changes in data over time and across groups or categories
- can be more useful than tables for understanding and analysis because of their ability to highlight trends, patterns and anomalies
- if well designed, can make a publication more attractive and may motivate people to read the whole document
- appeal more to a general audience than tables.
However, graphs can become cluttered if too much data is presented, can make it difficult to determine precise data values, and can be manipulated so that the message becomes distorted or erroneous. Graphs need to (Tufte 2001):
- show the data
- emphasise relationships between the data
- avoid distorting what the data have to say
- encourage comparison of different pieces of data
- balance accuracy, scientific standards and visual appeal
- be closely related to accompanying text and descriptions of the dataset.
The term ‘chart’ is often used interchangeably with ‘graph’, but ‘chart’ is a broader term that can encompass diagrams, maps, tables and graphs.
This section covers:
- Choosing the right graph for your data
- Types of graphs and plots
- Interactive and dynamic data visualisation
- Main conventions for graphs
- Functional design for graphs
- Things to avoid
- Preparing graphs for publication.
Download our quick guide for easy reference: What type of graph is best for my data? .