When your thesis arrives on the examiner’s desk, the first thing they will do is flick through it quickly before reading it properly (just like you do when you pick up a book for the first time).
The first things they will see – and your first chances to make an impression – will be the formatting and the figures, so it’s important to do them well.
I’ve written about the importance of formatting elsewhere, so here are my basic tips for designing figures.
Clarity is King
The aim of a figure is to convey information clearly and efficiently, and this must be foundation of any design decisions you make. There’s no point in making them eye-catching if you make them illegible in the process.
Design for print, not just screen
The figures have to look good when printed, not just on screen. Most modern computer screens are better than most printers, but even if you do have access to professional quality printing facilities, bear in mind that other people might need to print your thesis too (or that you might have to print your thesis in a hurry on deadline day using the first printer you can find).
How to write your PhD thesis: The secrets of academic writing
21st November 2018 2018
This is quite important when designing figures in colour- if you have a graph with several data series differentiated only by colour, the contrast may disappear when printed in greyscale using a black and white printer.
You could do a test-print to make sure it looks OK in black and white, or you could just design the figure that way from the start. On any printer, it will still be possible to differentiate between the data series in the example below.
Limiting yourself to black and white or grayscale forces you to think about clarity. If you need 20 different colours then perhaps the figure is too complicated.
It also forces you to be consistent in style. Colour does not always look better, and an inconsistent palette can look amateur.
Of course you can use colour if you really want to, but print the page in black and white and make sure it’s still clear.
Avoid JPEGs (except for photos)
JPEG is an excellent method of image compression for photos, but terrible for figures requiring sharp lines or text. Often there will be a dappling effect around lines or text making the figure look dirty.
Image compression is too big a topic to cover in detail here, but you should be aware that the file formats make a big difference to the final quality and different types are suitable for different purposes. PNG, for example, is more suitable for the web than for print.
I used EPS format for the figures in my thesis as it allowed true vector graphics (lines look sharp no matter how much you zoom in or resize the image), though this format can be a bit awkward to use. If you are using Microsoft Word, enhanced metafile (.emf) is a good option.
If you must use JPEG or PNG, use the highest quality compression you can when you save the image, but be careful not to create huge files.
Avoid converting images between formats, because each conversion degrades the image. Keep the original data, and then save in the format you intend to use from that original source.
Make sure the text is legible when printed
I often see figures with text so small it’s impossible to read, making them useless. In journal articles there is often a limited space for images, but in your thesis you can take as much space as you need. There’s nothing wrong with using the whole width of a page (or even a whole page) for a figure—if it’s important enough to be included in your thesis, it deserves as much space as necessary to make it legible.
Be careful when resizing though. Text might become distorted if you’re blowing up a low-resolution image, in which case you need to go back to the original and re-save as a higher-quality file.
Again, it needs to look good when printed, not just on your screen, so if in doubt do a test print to make sure it’s legible.
Avoid fancy backgrounds
If you’re saving a graph from Excel, you’ll have the option of a coloured (or even textured) background. Avoid this, it looks terrible.
I know you're probably busy right now...
Would you like to receive my top 7 articles to read in your own time? These are some of the most important principles I think every PhD student (or academic) should know. Enter your name and email and I'll send you one per day for the next 7 days.