Formatting tables and figures

It’s good practice to prepare your tables first (and discuss them with your advisor), even before starting to start writing your manuscript.

In general, tables need to be formatted yourself, e.g., in Word, Excel or Latex. Put differently, do not merely paste the output from your statistical program in the main text - these tables are often too elaborate, or do not efficiently combine the results of multiple models).

  • Seek inspiration from how tables look in published articles in good journals.
  • Make use of packages that format or combine output of your statistical analyses (e.g., stargazer in R).
  • The same holds for figures; the R package ggplot2 is recommended, but also R’s regular plotting functionality is great.
  • For more advanced tables, you can design them from scratch (e.g., in Word or Excel and manually paste numbers), or automate the entire process using tools such as TableFill.

Make yourself understood

  • Tables and figures should be understandable without reading the corresponding sections in the manuscript.
    • Use a clear title: easy to understand, and comprehensive (e.g., summarize the main point in the title (e.g., “X increases Y”), rather than using generic titles (“Results”).
    • Provide comprehensive table notes wherever necessary.
    • Use clear labels, e.g., for variables
  • Avoid colors, also in figures. Most readers will print your thesis in greyscale.
  • Note that you can also combine tables into larger ones, and thereby save space (e.g., tables with the same explanatory variables but different coefficients).
  • Generally speaking, tables and figures need to be understood without reading the main text.