Effective writing supervision for PhD candidates

Gesprek met promovendus

Writing is an essential part of the scientific process: it is the most common and important way for researchers to share their findings. Learning to write is often challenging for PhD candidates. As a supervisor, you play a pivotal role in supporting this process. This article is primarily written to support the effective supervision of PhD candidates in their writing process, but many of the situations and possible solutions discussed are also relevant when supervising bachelor’s and master’s theses. 

Challenges in Supervising PhD Writing

Many supervisors experience writing supervision as difficult and time‑consuming. Supervisors feel that their PhD candidates do not always process their feedback well enough, which can lead to frustration. Moreover, finding the right balance between providing guidance and stepping back is complex. This is because you, in your role as supervisor, often have multiple (conflicting) roles. As a co‑supervisor, for instance, your responsibility is to let the PhD candidate develop as an academic writer, which at times requires you to “let go” and take ownership of the text. This can conflict with your role as a co‑author, where you are expected to take a more active role. A co‑author, after all, contributes directly to the text.

PhD candidates, in turn, often report feeling anxious about the feedback they receive. They frequently do not know how to process it. Getting stuck in the writing process is often mentioned as a cause of delay, or even a reason to abandon a PhD trajectory altogether (Cassuto, 2013)

Therefore, effective supervision throughout the PhD trajectory is crucial (Johnston et al., 2016). Effective supervision means that you activate your PhD candidate. In this way, you facilitate that your PhD candidates learn to write with increasing independence, motivation, and confidence. Below are some of the most common challenging situations in supervising PhD students, and strategies for addressing them effectively.

Situation 1: The PhD candidate does not submit a draft

In this case, many supervisors simply extend the deadline. Yet, the underlying reason is that PhD candidates oftentimes are afraid to submit something because they know it is not yet “good enough.” This fear is reinforced by the expectation of large amounts of (conflicting) feedback (from multiple supervisors)

Suggestion: Identify the cause of the writing problem by asking open questions.

When a PhD candidate gets stuck, there is a tendency to immediately mote to solution mode (arrow 1 in the model below). However, the cause of the writing problem can lie in many different areas. Are personal factors at play, such as motivation issues described above? Does the PhD candidate know what is expected of them? Are they actually ready to write, or are there still tasks (e.g. analyses) that need to be completed before the writing can start? And does the PhD candidate know how to address the problems they are encountering (see also Suggestion 2)?

Avoid moving to solutions directly; instead, ask open questions. For example:

  • How did writing this text go for you?
  • How did the writing process go, which steps did you take?
  • What are you satisfied with, and what still needs attention?
Possible routes for solving writing problems image
Model for Supervision of writing and research

Possible routes for solving writing problems (from De Jong & Van Kruiningen (2022), p. 26). As supervisors we tend to offer ready-made solutions to a writing problem as quickly as possible (arrow 1). However, this is not always effective, and certainly not always efficient. Therefore, it is useful to take a detour. First, by (jointly) searching for the cause of the problem (arrow 2). Then by (jointly) looking for strategies the PhD candidate can use (arrow 3) to solve the writing problem (arrow 4). Even more efficient is to (jointly) search for strategies that enable candidates to evaluate their own work, so they can already identify problems independently (arrow 5) or articulate them clearly (arrow 6).

Situation 2: The text is chaotic

Often, the first reflex is to edit the text. But that takes a lot of time and is not always effective. Moreover, if you do not yet know what is causing the text to be chaotic (see Suggestion 1), you may end up doing the work for the PhD candidate rather than with them. As a result, the PhD candidate misses an important opportunity to learn how to produce a clearer, more coherent text independently.

Suggestion: Offer writing strategies that PhD candidates can apply independently

It is often assumed that PhD candidates know what writing strategies are, but this is by no means always the case. Or they know them in a general sense, but may never applied them at this scale or within this discipline. By offering writing strategies instead of ready‑made solutions, you encourage ownership and independence. Examples of writing strategies include:

  • "Circling around a question" by Heinze Oost which helps to analyse whether there is a valid research question and to check whether the basics (what, why, and how) of the research question are logical, clear and consistent (see pp. 51–104 of De Jong, 2017)
  • Creating an outline using questions and answers (see pp. 176–186 of De Jong, 2017)
  • Speed writing (see pp. 186–188 of De Jong, 2017)
  • Reverse outlining (see the UNC Writing Center for an explanation of reverse outlining)

Situation 3: Feedback is not processed or is misunderstood

Hurray, a text has been submitted at last. You have provided extensive feedback, but hardly any of it is processed. How frustrating! 

Suggestion: Match feedback to the writing phase

Discuss together why the feedback was not processed! Ask open questions:

  • What was the writing problem (see Suggestion 1)?
  • How much time did the PhD candidate spend processing the feedback?
  • What did the PhD candidate do to process the feedback?
  • Which strategies were used?

It is important to know that writing is a cognitively complex task (Flower & Hayes, 1980). It consists of multiple phases, including generating ideas, structuring and organising information, formulating sentences, and editing. If a PhD candidate is still working in the idea‑generation phase, giving feedback on grammar and style is usually not helpful. Segmentation of the writing process – working on one phase at a time – helps writers to maintain an overview and stay motivated

Situation 4: The text is not sufficiently academic

Many PhD students are told that their text is “not academic”. This makes them very insecure. How can you best support them?

Suggestion: Be explicit

Together, examine at what level the problem occurs: is it the research question, the content, the structure, or the style? Then consider which strategies your PhD candidate could use to make the text more academic. Analyse a strong and a weaker example text together, and discuss the differences. This helps making the meaning of “academic writing” more concrete

Conclusion

 Do you recognise the situations described above? Which suggestion would be most helpful in your day-to-day practice? And which suggestion will you try first to support PhD candidate(s) even more effectively in their writing process?

Want to know more?

There is much more to learn and know about (writing) supervision for PhD candidates. Read the sources listed below or get in touch for a professional consultation. For UU teaching staff, there are no costs involved. Onderwijsadvies en Training (Education Development & Training) also offers courses on PhD supervision.

References

Cassuto, L. (2013). PhD attrition: How much is too much? The Chronicle of Higher Education.http://chronicle.com/article/PhD-Attrition-How-Much-Is/140045.

De Jong, J. & van Kruiningen, J. (2022). Geef ze het nakijken: over activerende feedback en schrijfbegeleiding in het hoger onderwijs. Boom.

De Jong, J. (2017). Effective strategies for academic writing. Uitgeverij Coutinho.

Flower, L., & Hayes, J. R. (1981). A Cognitive Process Theory of Writing. College Composition and Communication, 32(4), 365–387. https://doi.org/10.2307/356600

Johnston, L., Sampson, K., Comer, K., & Broght, E. (2016). Using the doctoral experience survey data to support developments in postgraduate supervision and support. International Journal of Doctoral Studies, 11, 185-203. https://doi.org/10.28945/3505