Doctoral Project Failures May Start With Supervisors, New Risk Framework Argues

When a doctoral project begins to unravel, the explanation that usually follows is a familiar one: the student lacked the ability, the motivation, or the resilience to finish. A new conceptual study published in the journal Higher Education challenges that reflex, arguing that many PhD projects become unstable not because of isolated student shortcomings but because of predictable vulnerabilities built into the design of the research itself — vulnerabilities that supervisors are uniquely positioned to create, amplify, or prevent. The article, authored by Catherine M. Worsley of the University of Pretoria, proposes a supervisory risk framework that reframes doctoral instability as a systemic and developmental problem rather than a personal deficit.

The framework identifies five interrelated risk domains that emerge during early candidature: research question scope, methodological alignment, timeline design, communication structures, and expectation alignment. Crucially, these are not treated as discrete problems that surface later in a student’s journey, but as interacting elements of supervisory design whose consequences accumulate quietly over months and years. By the time difficulties are recognised, the article argues, substantial time, resources, and emotional investment have often already been lost. The aim is to shift attention from reactive remediation toward preventive supervisory design, embedded within the institutional performance environments in which modern doctoral education actually takes place.

The first domain, research question scope, is distilled into a deceptively simple supervisory question: is the project asking something that can realistically be addressed within a doctorate? A question can be intellectually compelling and conceptually ambitious yet still exceed what can be achieved within the available time, resources, and supervisory capacity. Early warning signs include questions that attempt to address too many factors simultaneously, projects that bundle multiple distinct aims into a single study, and assumptions about data availability or practical feasibility that have never been adequately tested. Worsley notes that supervisors themselves may inadvertently encourage overambitious questions in pursuit of originality or publication potential, or delay narrowing the scope in the belief that refinement will happen naturally as the project matures. Without explicit boundaries, projects tend to expand rather than converge, increasing the risk of stalled progress and student frustration.

The cumulative consequences of scope risk are illustrated through a typical pattern: a doctoral student proposes to investigate multiple aspects of a broad problem within one project, simultaneously examining underlying mechanisms, comparative methodologies, and applied outcomes. At proposal stage the project appears coherent and academically strong. Once work begins, however, each component demands distinct methodological expertise, extended optimisation, or additional data sources, and supervisory discussions drift into incremental troubleshooting rather than strategic re-focusing. By the time substantial scope reduction becomes necessary, the student may have invested significant time in work that cannot be integrated into a coherent thesis narrative. Similar dynamics arise in humanities contexts when projects attempt to engage multiple theoretical frameworks at once, producing diffuse analytical focus. The proposed remedy is early and deliberate: requiring students to articulate their question in a single tightly defined sentence, distinguishing primary from secondary aims, conducting feasibility checks against confirmed resources, and introducing a formal ‘scope freeze’ at the proposal or early candidature stage.

The second domain, methodological alignment, asks whether the chosen approach is genuinely feasible for answering the research question. Designs that look appropriate on paper can prove difficult to implement once practical demands are considered, particularly when supervisors and students try to maximise impact by incorporating multiple techniques, large datasets, or unfamiliar approaches. Early progress can be seductive: initial data collection, pilot experiments, or preliminary analyses create a temporary sense of feasibility that masks the exponential increase in complexity as the project scales. When setbacks arrive — failed experiments, inaccessible datasets, limited archival access, unavailable fieldwork sites — students may respond by broadening their methods rather than refining them, widening the feasibility gap and destabilising the proposed timeline.

That timeline instability forms the third domain. Timelines are frequently constructed in good faith but on optimistic assumptions about methodological readiness, administrative efficiency, and uninterrupted progress — assumptions that rarely survive contact with real research. Administrative processes such as ethics approvals, procurement delays, and site access are treated as minor formalities rather than substantial time variables, while plans assume linear progress despite research typically unfolding in non-linear cycles of refinement and revision. The article stresses that these compressed planning practices are reinforced by institutional performance cultures that prioritise timely completion metrics and throughput targets, meaning ambitious timelines may be implicitly rewarded even when they underestimate research’s inherent uncertainties. The consequences go beyond missed deadlines: when early delays occur, students may narrow their analyses prematurely, avoid necessary methodological adjustments, or overextend themselves to catch up, producing projects that are practically complete but intellectually constrained. Worsley recommends retrospectively constructing timelines from key submission deadlines with explicit buffer periods, and supervisors openly discussing prior project delays to normalise iterative development.

Communication structures, the fourth domain, concern the agreed processes through which information is exchanged between supervisor and candidate. Breakdowns here rarely stem from overt conflict; they develop gradually through unexamined assumptions about reporting, independence, and help-seeking. Early indicators include unstructured meetings, reduced disclosure of uncertainty, delayed circulation of drafts, and uncertainty about when difficulties should be raised. A recurring pattern sees students withholding problems to demonstrate competency, interpreting uncertainty as personal deficiency, while supervisors interpret reduced communication as a sign of productive independence. In the absence of explicit reporting structures, silence can be misread as progress. Mismatches over feedback expectations compound the problem: students may assume drafts should be near-final before review, while supervisors expect iterative, exploratory drafts and read delayed submissions as disengagement. These dynamics are intensified in performance-oriented settings, where pressure to present smooth progress narratives narrows communicative space precisely when open dialogue would be most protective. The remedy is formalisation: structured meeting schedules, brief written progress summaries, explicit normalisation of methodological setbacks, and clear distinctions between independence and isolation.

The fifth domain, expectation alignment, underpins all the others. It concerns the implicit assumptions shaping how supervisory roles, project priorities, progress, and outcomes are understood. Students may expect close guidance with frequent direction, while supervisors view doctoral research as an apprenticeship in increasing autonomy; supervisors may prioritise publication-quality outputs while students prioritise timely completion or skill development. When these orientations remain implicit, they generate subtle tensions that shape decision-making throughout the project — overly ambitious scoping driven by publication aspirations, compressed timelines motivated by completion pressures, and communication breakdowns rooted in differing views of independence. Structured expectation setting through written supervision agreements, explicit discussion of what constitutes satisfactory progress, and periodic revisiting of project goals can make these assumptions visible and revisable, reducing the risk that students misinterpret ordinary challenges as personal inadequacy.

What gives the framework its broader significance is its positioning of supervision as a meso-level mechanism — the intermediate layer through which institutional conditions actually reach individual doctoral projects. Institutional expectations do not shape PhD projects directly; they are mediated through supervisory decisions about scope, timelines, methodological approaches, milestones, and expected outputs. When universities emphasise timely completion and research productivity, those pressures flow through supervisors into the practical conditions under which projects develop, including how readily emerging problems are recognised and addressed. This reframing moves doctoral instability out of the individualised deficit paradigm and into a systemic design problem, connecting project-level design, supervisory practice, and institutional conditions in a single analytical lens.

The author is careful to acknowledge the framework’s limits. The analysis is conceptual rather than empirical, derived from supervisory experience and existing literature rather than systematic investigation, and should be read as a stimulus for reflection and preventive action rather than a validated predictive tool. It may not capture disciplinary variation in full, and it does not fully account for institutional constraints on supervisors’ capacity, nor for individual agency, motivation, financial pressures, employment alongside study, or caregiving responsibilities that also shape candidature trajectories. Nor does the approach eliminate the uncertainty inherent in research or remove student responsibility. Future research, Worsley suggests, should test the framework through longitudinal, cross-disciplinary, and mixed-methods studies examining whether the proposed risk domains genuinely predict instability, completion delays, or attrition, and how institutional policies and workload structures enable or constrain preventive supervisory practice. The framework does not remove student agency and responsibility; rather, it positions supervision as an intentional process of project design and ongoing recalibration, supporting doctoral projects better able to accommodate uncertainty while reducing preventable escalation and promoting more sustainable postgraduate development.

Subject of Research: A supervisory risk framework explaining doctoral research project instability through interacting design vulnerabilities in supervision.

Article Title: Reframing doctoral research instability through a supervisory risk lens

Article References: Worsley, C. M. (2026). Reframing doctoral research instability through a supervisory risk lens. Higher Education. https://doi.org/10.1007/s10734-026-01779-y

Image Credits: AI Generated

DOI: 10.1007/s10734-026-01779-y

Keywords: doctoral supervision, doctoral education, PhD completion, research design, supervisory risk, higher education, project instability, expectation alignment, timeline design, academic socialisation, institutional performance, preventive supervision