Deskins Tips on Tables and Charts

Making Good Figures and Tables

The heart of any scientific document are the figures and tables, which convey the most important information to the reader or viewer. Most of the time you'll probably use Microsoft Office or other software to make your figures and tables. The default options of many programs however just look wrong for a professional, scientific document. If you want others to take your work seriously, you need to look serious. Below are some examples of amateur and professional plots and tables. Other examples can be found here.

Every figure or table should also have a caption explaining it. Everything the reader needs to understand the figure/table should be in the caption. The reader should not need to read the text to understand the figure/table.

Note: these images are all grainy because of low-resolution. Ignore this. You should use at least 300 dpi in your documents. Also, for journal articles the typical column width is about 3.5 inches, so you should make your your figures look good in this size. Learn more at Image Quality.

Plots

Good plots have axes labeled, no title, no legend (unless you have more than one data set), and no surrounding box. Make sure units are in axes titles. Use black and white for a single set of data. If you are plotting multiple data sets you can use different colors, like red, green, etc. But don't use fancy or non-traditional colors.

Tables

Tables should be simple, black and white. The label row (top row) should have top and bottom lines. The final row should have a bottom line. Labels with units should be at the top of every column. The table should be an appropriate width (not too wide).

Pictures

Some of your figures will be pictures. They should be simple and understandable. If the picture is confusing you should remake it. Use a white background. Label parts if necessary. Getting good pictures can be difficult, and you may need to ask others if they understand what you are trying to convey.

Captions

Again, the captions should be self-contained, or have everything that is needed to understand the figure or table.

Amateur Data collected for the calorimetry experiments.

Professional Energies (kJ/mol) and temperatures (K) for several calorimetry experiments. All data was collected using 500 L of 99.9 wt. % ethanol mixture.