Writing together

From Bayesian Behavior Lab

Jump to: navigation, search


Let's write together

I find that the best way to both get papers published in a good form and to teach my students/postdocs how to write is to write together. I learned that from Daniel Wolpert who is a great master of writing. I have the students write the abstract (early in the project) or a first draft (later) all by themselves. But then instead of me editing over it by myself I edit it with them also looking at my screen. I usually only do this once per paper but this run-through is super super slow.

Why I believe that it is a good idea

  • I cannot write something that is wrong. The student sees me and immediately challenges me on it.
  • I cannot say something that cannot be understood. They will spot that rapidly.
  • They see how I approach writing. They see the various approaches I use to make a paper better. Also see Paper Writing 101. I find that writing together accelerates their learning curve.
  • They get careful feedback on their writing weaknesses - and get to discuss to see the rationale why I make changes. They also get to see the weaknesses in my way of writing.
  • By having to express why I edit the way I do I become far more mindful of the reasons.
  • By having constant criticism from all my students I improve my writing ability.

Why other people might not write together

  • We might waste their time. But really, I think this time is well worth it. It leads to better papers. And an understanding of the process.
  • We might show that we have weaknesses in writing ourselves. If anything this should be a reason to do it. I am rather worried that my students will respect me too much and hold back relevant criticism.
  • We might waste our own time. If they ask questions then we have to answer them. I neither believe that it really wastes our time nor that it would be a bad time if it did cost us some time.


  • I have students fill in the references and do not finalize it during our joint edit. They have a different endnote file and in any case, I find that they are very reliable on this.
  • I have them finalize the figures. Doing this well takes too long for a joint session.
  • I sometimes have them figure out the microstructure of the text. How one sentence feeds into the next can often be improved iteratively in a good way.
This page was last modified on 14 February 2014, at 16:37. This page has been accessed 544,282 times.