Graphs can quickly convey relationships and trends between and among data in ways that a data table or text narrative struggles to achieve.
Each type of graph has a purpose that it is best suited for.
Some have purposes they should never be used for.
Editors who know these principles can more effectively ensure that the graphs suit their intended purpose and fit scientific and mathematical standards for data communication.
This is a round-up of advice and resources explaining the best practices (and some pitfalls) for data presented in graph form.
Pitfalls of Graphs
- Exaggerated scale (affects perception of line slope / change over time)
- Scale doesn’t start at zero (affects perception of scale of change)
- Type doesn’t suit data (implies relationships that don’t exist)
Types of Graphs and Their Best Uses
Editing by Design, Jan White, pp. 159–162, 168.
Statistics Canada’s guidelines for graphs
See chapters 16 & 17 in Geoff Hart’s Writing for Science Journals