Dos and Don'ts when Writing a Paper

There are some dos and don'ts when writing a scientific paper that heavily improve the reading. In the sequel, I provide some hints I have learned as author and reviewer in the past. If following list lacks some important hints, contact me and I will add them here.

Improving the Readability of a Scientific Paper

  • Try to reduce following modal verbs:
    • can, could, must, have to, should, would
  • Cite papers, call rules, give hints, or move derivations to the appendix instead of using following sentences:
    • "It can be easily shown that..." 
    • "After some computation..." 
    • "It is straightforward (to show that...)"
  • As a rough guideline use 
    • more than 10 references for conferences papers,
    • more than 30 references for journals papers,
    • other references than only yours.
  • In each sentence, present only one idea. Write condensed.
  • Use LaTeX instead of Word. Use following LaTeX's packages:
    •  TikZ and PgfPlots for figures and Matlab plots.
    • SIunitX for (SI) units
    • MathTools for typesetting of equations
  • Use captions that explain the figures/tables.
  • Mathematical symbols are roman, if they are descriptive 
    • Superscript and subscript are italic, if they are variables or indices, else roman:  $x_k, k \in \mathbb{Z}$ vs.  $x^{\mathrm{MSE}}$, or $x^{\mathrm{T}}$ (transposed) 
    • Functions are roman, if they are descriptive, else italic: $f(x)$ vs. $\sin(x)$ (sine) or $\mathrm{E}(x)$ (expectation) or $x^{\mathrm{T}}$ (transposed)
    • Units use roman fonts (see below). 
  • Labels of axes in function plots:
  • Use tables for a long list of simulation and measurement settings. 
  • Tables should be set as described by http://mirror.klaus-uwe.me/ctan/macros/latex/contrib/booktabs/booktabs.pdf:
    • Never, ever use vertical rules.
    • Never use double rules. 
    • Put the units in the column heading (not in the body of the table).
  • At the end of the paper you don't need a second abstract. Draw conclusions instead.
  • Use less than 5 abbreviations. Do not use them in the paper's title. 
  • Divide the introduction into Motivation, State of the Art,  Contribution(s) and Outline. 
  • Simplify notations. Use sub-/superscripts instead of different symbols, but use as few as possible. 
  • Motivate your work by means of an application. 
  • The  "delta-Dirac function" is either a distribution or an indicator function, but this should be specified.
  • Do not use \jmath and \imath for $\sqrt{-1}$ in LaTeX.
  • Use hyphens: especially compound modifiers.

Acknowledgements: Paul-Jürgen Wagner's comments helped me to improve and extend the list. 

Further Reading