I have supervised five doctoral candidates, four of them now completed, as principal supervisor for two and co-supervisor for three. All five held competitive scholarships, three under the Research Training Program and two from the China Scholarship Council. That is a small number, and it is worth saying at the outset that co-supervision means someone else carried more of the load.
Where they went
Kun Wang completed in 2023 under my principal supervision and is now at Ningbo University. Her thesis work on social sustainability in aged care and hospital partnerships became a body of joint papers that is still running, and it is the part of my own record I would be least willing to give up.
Noah Mwelu completed at the University of Newcastle in 2020, where I co-supervised, and is now at Makerere University Business School in Kampala. His work on compliance in Uganda's public road construction produced three papers, one of which was later cited by Tanzania's Public Procurement Regulatory Authority in a report on procurement practice. A thesis reaching a neighbouring government's regulator is a better outcome than most of us get.
Xiaohang Xu completed in 2022 and now works in district government in Shenyang. Her work on opportunism in partnership projects produced a paper arguing that opportunistic behaviour does not always reduce stakeholder satisfaction, which is a harder claim to publish than the conventional one and took several years to place.
Juanwen Liu completed in 2024 and is now with PetroChina International. Her work on value capture in social infrastructure fed into a paper that took a merit award at CRIOCM in 2024.
Ibrahim Dani is the candidate I currently supervise as principal supervisor, working on artificial intelligence and project management. His first paper, a narrative review of project management in industry 5.0, was published in Project Management Journal and became one of the journal's most read and most trending articles in 2026. Publishing well in a first paper does not guarantee a good thesis, but it removes the question that occupies most candidates for the first two years.
The candidate is the project manager
I teach project management, and I think the discipline has something to say about doctorates that supervisors mostly avoid saying. A doctorate is a project with a fixed schedule, a constrained budget, a single deliverable, high uncertainty about scope, and a project manager who has never run one before. That project manager is the candidate. It is not me.
The default model runs the other way. The supervisor sets direction, the candidate produces work in response, and progress depends on how often the supervisor initiates. That arrangement can produce a thesis, and I have seen it produce good ones, but it fails badly when the supervisor is busy, when the fit is poor, or when the candidate is capable of more than the supervisor is asking for. It also produces graduates who have not learnt to run a research program, which is the actual thing a doctorate is meant to certify.
A supervisor is a resource with limited availability, particular expertise and known blind spots. Learning to use one is part of the training.
I say this to candidates early and I mean it operationally. Set the agenda for the meeting and send it beforehand. Send work in advance with a specific question attached, because a draft sent with "any thoughts?" will produce a slow and general answer. Say which kind of feedback you want, structure or wording, and do not ask for both in the same pass. Keep your own record of what was decided, because I will forget what I said three months ago and will occasionally contradict it. If you have a panel, use the members differently rather than asking all of them the same question. And chase: a supervisor who has not replied in two weeks has forgotten, not decided.
What I owe in return has to be concrete or the argument becomes an excuse. I commit to a turnaround time and say so when I will miss it. I tell candidates when work is not yet good enough at the point I think it, rather than at the point it is submitted. And I say plainly whether I think a chapter is publishable, since a vague answer there costs a candidate a year.
Sprints rather than a marathon
The second half of the argument is about how the work is broken up. A thesis treated as one deliverable due in three or four years produces long stretches in which nothing is finished and nothing is visible, and by the time a problem surfaces at a progress assessment it has usually been there for a year. Treated as a series of short cycles, each ending in something a person other than the candidate can read, the same project becomes correctable while there is still time to correct it.
This is not a metaphor I invented for research; it is the ordinary reason iterative delivery exists, applied to a project that happens to be a thesis. I organised a workshop on it, Agile Your HDR: Turning a Marathon into Manageable Sprints, because the pattern I keep seeing in milestone assessments is not weak candidates but long undivided stretches of work with no external check inside them.
Where this argument fails
Ownership language can be used by supervisors as cover. A supervisor who tells a candidate to take ownership and then does not read drafts has not transferred responsibility; they have abandoned it and given it a better name. The obligation does not reduce because the candidate has accepted more of it.
It also assumes a capability that most candidates do not have on arrival. Someone in their first year, often working in a second language and in an unfamiliar system, cannot manage a resource they have not yet learnt to read. Early on I do more of the driving, and the transfer is the thing being taught rather than the condition of entry.
The agile framing has a real limit as well. Some problems only give way to long uninterrupted attention, and a candidate who has been trained to produce something reviewable every few weeks can end up avoiding exactly that kind of problem. Short cycles suit data collection, analysis and writing. They suit the year spent working out what the question actually is much less well.
And the power asymmetry does not disappear because I have described it. Telling a candidate to manage their supervisor as a resource is easy advice to give from my side of the table. The candidate cannot replace the resource, cannot escalate without cost, and depends on it for examination, references and often a visa. Any version of this argument that ignores that is dishonest.
HDR coordination and examination
Since 2024 I have been the School Higher Degree by Research Coordinator, which puts me in Stage 1, Stage 2 and Stage 3 assessments across the school rather than only in my own candidatures. That is where most of what I think about supervision comes from: the recurring problems are project definition, feasibility, writing quality, supervisory alignment and timely progression, and they repeat across disciplines and supervisors. Since the same year I have served on the Faculty Research Degree Committee, assessing applications and reviewing examination reports, and from 2025 on the Graduate Research Students' Appeals Committee, which is where the failures of all of the above eventually arrive.
I have examined doctoral theses as an external examiner at Lingnan University, City University of Hong Kong, the University of Sydney, Western Sydney University, the University of Newcastle, Curtin University and RMIT University. Examining is the fastest correction available to a supervisor's judgement, because it shows what a thesis looks like to someone with no investment in it.
From Aug 2026 I hold a two-year appointment as Associate Professor of Graduate Research in the UTS Graduate Research School, designing and delivering supervisor development across the university. Alongside that I have run sessions for candidates and supervisors on conducting a literature review, on improving the chances of paper acceptance from an editor's perspective, and on topic selection and writing, and I co-wrote the guidance paper on writing literature review articles for Project Management Journal, which was one of that journal's most read articles across 2025 and 2026.
A note on where candidates come from: at least one current doctoral candidate in the school decided on research after taking the postgraduate research subject described on the teaching page, and won a competitive scholarship on the strength of it.