Learning how to write a research paper is difficult and takes time, yet it is an integral part of a PhD in STEM. Teaching your students and postdocs how to write good papers is an essential part of being a good supervisor and mentor.

Loose keys from a rainbow coloured computer keyboard

This post is about a particular aspect of writing papers that I noticed many of my students struggle with: Focusing on a specific question. Related to that is the issue of creating a narrative. The source of this struggle is complex and usually more a reflection on how we teach STEM at Universities rather than an individual student’s ability.

In my experience, there are two main sources. First, we rarely teach students to formulate questions and then think of specific experiments and analyses to address them. In most laboratories and tutorial classes, we give students a set of questions and pre-designed experiments or data sets. This is ironic, given that asking questions and designing experiments is the essence of science. Second, in STEM undergraduate courses, students are rarely taught to think of a paper in terms of a narrative or a storyline. We teach students the generic recipe of Introduction, Methods, Results and Discussions and forget to teach them that there still needs to be an overall storyline or focus.

In my field, the difficulty of finding a focus is made worse because of the nature of the raw data our in-silico experiments produce. The data from simulations of biological or chemical systems can be used to calculate a vast array of properties. The analysis you can do seems endless. This is made worse because, in my experience, many students think that the more data or analysis they present in a paper, the better. Unfortunately, this is rarely the case. Instead, when students lose focus on the specific questions that they want to answer, the result is a paper where the signal is lost in the noise. Or, as my former postdoc supervisor and mentor says, “Analysis for the sake of analysis is rarely a good idea.”

I have recently started using the following approach to help students plan their analyses and paper. Once a student has collected data, I asked them to write 1 to 2 pages to answer the following questions.

  • What is the overall question you want to answer in this paper? Or, what is the hypothesis you want to test in this paper?
  • How does this question or hypothesis relate to what we know about this system or problem?
  • What is the minimum data you need to answer your question or test your hypothesis?
  • What specific questions do you need to ask to address your overall questions or hypothesis?
  • For each of the specific questions, what data and what analysis will you use to answer it?
  • How is this data or analysis best presented?
  • How does each datum or analysis relate to the overall question or hypothesis?
  • What does the data tell you with respect to the question or hypothesis?

Based on the answers to these questions, I get the student to map out the Results and Discussion sections. I ask them to produce an outline with the order of all plots and figures and a storyline through which they build their arguments towards answering the overall question. For each plot or figure, the student needs to tell me why it is required to make their argument.

P.S. I understand that sometimes the storyline and argument only emerge as you write. But in my experience, most students need the above support structure before they start writing. This does not mean the paper cannot evolve as you write and edit.

P.S. The idea of writing a scientific paper as a story has its advocates and opponents. Below are two interesting blog posts related to this. While I don’t go as far as thinking of my paper as having characters, settings, actions, climax and resolution, I still find it helpful to think of my paper as having a storyline. I agree with Thomas Basbøll that “It is more efficient to think of your paper as series of claims to be supported, elaborated or defended.” I also like Aimee Edgeworth’s interpretation of a storyline in the context of a research manuscript.

More Resources

The Guiding Principle in Scientific Writing

5 Things That Help Me Write

The Write Rules   

Want to live on the Edge?


Join the conversation

Your email address will not be published. Required fields are marked *

Saving subscription status...


You May Also Like