(Adapted from an answer I gave several years ago to this question on Quora)
I’m assuming you are actively writing. If not, you must start there.
Here are some things I have tried to follow to improve my scientific writing:
- Understand the basic parts of a scientific paper. This might sound trivial, but it is the essential first step. Know what should be included in the abstract, intro, methods, etc.
- Know the relevant background material very well. This will help you formulate and communicate your work in the context of the field. As a result, it will sound (and be) more refined. You may find my system for literature reviews useful: https://benjaminrigsby.com/a-systematic-guide-to-literature-reviews/
- Design your experiment (hypothesis, methods, etc.) as carefully as you can. This relates back to the previous point about knowing the background. Don’t put yourself in an unnecessary bind where your work is not clearly novel and useful.
- Make good use of figures and spend time to make them look nice and professional. If someone reads or even looks through your paper, they’re definitely going to see them. Don’t let your figures be a weak point of your paper. As a rule of thumb, if the figures are weak, the paper is weak.
- Have a good and experienced writer review your paper. Pay attention to what they say, fix your current draft, and remember it the next time you write.
- Pay attention to common technical words in your field and use them. This will improve the “technical” aspect of your paper. Also watch this superb talk on how to write good papers at least twice: https://www.youtube.com/watch?v=vtIzMaLkCaM&t=1s.
- Print your paper out and hand write your revisions. This forces you to concentrate on finding the mistakes rather than getting sidetracked half way through with fixing them. It also helps to see the paper in a little bit of a different perspective: a tangible piece of work.
Expect to have many revisions. That just means your paper is improving and you’re learning.
Happy writing and good luck!