The do’s and don’ts of data visualisation
The following techniques can be generally applied to a variety of visualisation techniques. When creating a figure, consult this list to make sure you’ve covered the basics.
For figures to convey your messages most efficiently, they can’t try and explain too many ideas at once. Make sure you have a specific message or purpose in mind for your figure – this makes it easier to cut down on any redundant aspects. Consider how an observer without your background knowledge would approach your figure for the first time. Be aware that this is more difficult than it seems. Scientists are often afraid that by making a simple figure, their work might seem basic or uncomplete. However, conveying your message correctly and accurately to your audience should always be your first priority.
As word processing software keeps our text looking sharp, we need high quality figures to go along side it. There are three common formats that our images are saved in. JPEG is best for saving photographs but should be avoided for graphics (such as lines, drawings, typography). For scientific figures, PNG is an ideal format as the details of the original image or conserved after compression. PNG also supports transparent background layers. When creating images using graphics software, the best format is SVG which scales endlessly while conserving sharpness.
Whether they are multiple figures for an essay or presentation, the look and feel should be consistent throughout. Make sure the same font and sizing can be seen across different figures. If one colour convention is used, make sure to remain consistent.
Though the figures should look the same, there should be diversity in what they are trying to explain. Avoid creating figures that express the same conclusion. This adds redundancy, taking up space in your work without helping to explain anything new. Give each of your figures a purpose, and make sure you employ them to tell your whole story.