A visualisation is composed out of different building blocks: there is text (like a title, labels and annotations), there is the data itself (represented by bars, lines, points or other geometrical elements) and there are guides for interpreting and reading of data (like axes, grid lines and colour legends).
The design of text elements is covered in the Typography and the design of text elements module. The design of data elements is a very broad topic and depends to a large degree on the chart type. This is covered in another training in this training program (Training 5. A deep dive into chart types). So this module covers the design of guides: supporting elements to aid the reading of data values.
But let’s start with the canvas charts are built on top of: the chart background. Usually the chart background is white, which makes it easy to guarantee that all elements of a chart have enough contrast with the background to be read easily, even the smaller elements and elements in muted colours like grey.
If you use a non-white background, make sure to check the contrast between the background and the lighter elements on your chart (see the Colour: the basics module for some tools you can use for that).
When embedding a visualisation in a publication, it doesn’t need a frame around it: this creates a “boxy” design. Charts without a border blend in better.
Source: Maarten Lambrechts, CC BY 4.0
The main part of the chart background is the plotting area. The plotting area is the plane enclosed between the x and y axes and is the area on which the data is plotted. Sometimes the plotting area is given a background colour. In that case, again, enough contrast must be ensured between the plotting area colour and the data plotted on it.
The default theme of the ggplot2 package in R applies a grey colour to the plotting area. Source: Maarten Lambrechts, CC BY 4.0
To give the reader some guides to interpret the data and read data values in a chart, you can provide axes, grids and legends.
Axes let readers assess the position and/or size of data marks. If a chart uses cartesian coordinates, the x axis is the horizontal axis and the y axis is the vertical axis.
The building blocks of an axis are the ticks, with their associated tick marks and axis labels, and an optional axis line. Next to the tick marks and labels, grid lines can also be attached to the axis ticks.
Source: Maarten Lambrechts, CC BY 4.0
Notice that the y axis of this chart does not have an axis line. Many data visualisation tools draw axis lines by default. But in many cases, these axis lines do not serve any real purpose, and they can be left out without any problem. This is especially true for bar charts, from which axis lines can be removed without any consequences.
A comparison of a bar chart with and without axis line. Source: Maarten Lambrechts, CC BY 4.0
Axis labels, however, are indispensable to interpret the data on a visualisation. Without axis labels, a visualisation is meaningless.
Source: Maarten Lambrechts, CC BY 4.0
The only exception to this are visualisations where enough data points are labelled so that axis labels are not needed.