Imagine a university adopting a doctoral supervision framework developed thousands of kilometres away. The policies are carefully written, candidature milestones are clearly defined, and supervisors receive formal training. However, the model assumes reliable internet access, regular attendance on campus, comfort in challenging academic authority, and a shared understanding of what knowledge is and how it should be produced.

What happens when those assumptions do not hold?

As higher degree research expands across institutions, countries and modes of study, we need to ask not only what good doctoral supervision looks like, but also where, for whom and under what conditions it works.

Doctoral education has changed

Doctoral education has undergone significant transformation. Candidate numbers have increased, and doctoral cohorts have become more diverse in their cultural, linguistic, disciplinary and professional backgrounds. Candidates may study on campus, online or through hybrid arrangements. They may be supervised by one academic or by a team spanning disciplines, institutions and countries.

The purpose of the doctorate has also broadened. Candidates are still expected to contribute original knowledge, but they are increasingly expected to develop transferable skills and prepare for careers within and beyond universities.

Hammond et al. (2010) describe four broad developments behind these changes: the formalisation of doctoral education, the growth and diversification of candidate populations, the diversification of modes of study, and the diversification of the doctorate’s purposes. Taylor (2023) similarly argues that doctoral education has been reshaped by substantial growth, regulation, structural change and increasing diversity.

These developments have moved supervision beyond its traditional master–apprentice origins. Historically, supervision was often treated as a private relationship between one academic and one candidate, conducted with limited institutional involvement. Today, universities are more likely to regulate supervision through policies, progress milestones, codes of practice, supervisor registration and professional development.

This greater structure and accountability can protect candidates and improve the quality of their research experience. However, standardisation can also conceal assumptions about how supervision should work.

What assumptions sit inside “good supervision”?

Many principles of good supervision appear universal. Supervisors should communicate clearly, provide constructive feedback, support candidate independence and create opportunities for intellectual development. Yet the ways these principles are understood and enacted can differ considerably between contexts.

Consider the expectation that a capable doctoral candidate should openly question or challenge their supervisor. In some educational and cultural contexts, this may be interpreted as evidence of critical thinking. In others, challenging a senior academic may be considered disrespectful or inconsistent with established relationships of authority. Power dynamics in the supervisory relationship remains a taboo topic and shape the conditions under which candidates can speak, question authority and exercise scholarly agency (3).

Similarly, independence is often understood as a candidate’s ability to work autonomously. However, this definition may not reflect collective, relational or community-oriented approaches to research. A candidate may understand responsibility for their research as extending beyond the individual to their family, professional community, cultural group or place.

Assumptions may also be made about what counts as legitimate knowledge. Established doctoral conventions can privilege written, individual and disciplinary forms of knowledge production. Indigenous, local, oral, relational or community-held knowledges may then be treated as additions to conventional research rather than as knowledge traditions in their own right.

The material conditions of supervision also vary. Access to reliable technology, specialist expertise, research infrastructure, professional networks and experienced supervisors cannot be assumed. Geography, institutional resources, academic workloads and access to research communities all shape what candidates and supervisors can realistically achieve.

Research supervisory practices must therefore be both conceptually and contextually relevant. They need a sound evidence base, but they must also respond to the particular cultural, social and institutional environments in which supervision occurs (4).

Context is part of supervision, not merely its background

It can be tempting to view context as a complication surrounding the central supervisor–candidate relationship. Yet supervision does not occur in isolation. It is shaped by institutional policies, cultural expectations, disciplinary traditions, geography, technology, funding and wider social relationships.

A practice may therefore be effective in one setting but less appropriate in another. Even within the same university, the needs of a laboratory-based candidate studying full-time on campus may differ greatly from those of a part-time candidate conducting community-based research from a regional location.

This does not mean every setting requires entirely different standards. Shared principles remain important. Candidates should receive ethical, respectful and intellectually rigorous supervision. Responsibilities should be clear, progress should be supported, and mechanisms should exist for addressing problems.

Contextualising supervision does not mean abandoning or lowering quality. It means examining whether our definitions of quality contain assumptions that have not been made visible.

Three questions for institutions and supervisors

When developing supervisory frameworks, policies and professional learning, we might begin with three questions.

What assumptions are built into our model?

Does the model assume candidates can attend campus regularly, readily challenge their supervisors, access specialist support or prioritise individual academic goals over other responsibilities?

Whose knowledge and communication practices are recognised?

Are multilingual, Indigenous, local, oral and community-based knowledge practices recognised as legitimate? Or are they expected to fit around conventions that remain unquestioned?

Who has helped define good supervision?

Have candidates, supervisors and communities from the relevant contexts contributed to the framework? Or has an existing model simply been transferred from elsewhere?

Best practice for whom?

The growth and diversification of doctoral education make shared guidance more necessary, not less. They also make it increasingly difficult to assume that one model of supervision will work everywhere.

Doctoral supervision does not travel unchanged. Every framework is interpreted through relationships, histories, cultures, institutions and places. The next time we describe an approach as “best practice,” we might pause and ask: best for whom, in which context, and determined by whom?

References

Banner image generated by ChatGPT