Working with your supervisor goes better when both of you know the rules, and most universities write them down in a code of practice: how often you meet, how quickly written work comes back, who keeps the record, and where to go when something goes wrong. This guide reads nine of them and sets the rules side by side.
It is the supervisor guide of our how to write a dissertation series, published by EdCitation. We read each rule below in the university's own pages on 26 September 2026: Loughborough, Imperial, Glasgow, Edinburgh and Bristol in the UK, Stanford and Michigan in the US, Toronto and Melbourne. Most of these codes govern research degrees; for an undergraduate or taught master's dissertation, your programme handbook answers the same questions, and Loughborough requires it to.
EdCitation never writes any part of a dissertation. It takes the mechanical checks off the supervisor's desk: Verify references checks a draft's reference list, free, before it is sent, so the meeting is spent on the argument.
What does a supervisor owe you, and what do you owe them?
A supervisor owes you guidance, availability, timely feedback and a warning when your work falls short; you owe them regular contact, written work on time and ownership of the project. Glasgow's code puts the balance in one phrase: the supervisory team gives "guidance, but not leadership" (University of Glasgow, 2025).
The supervisor's side
The University of Glasgow (2025) lists what every supervisory team takes on: being available for advice, asking for written work and returning it "with constructive criticism and in reasonable time", advising on completion dates, saying as soon as progress falls short, and keeping an agreed written record of any meeting where concerns are discussed. Melbourne's graduate research hub adds detailed feedback on thesis drafts, an agreed schedule of individual meetings, and reasonable access between them (University of Melbourne, n.d.).
Your side
Glasgow expects you to send written work regularly and early enough to be read, to keep in contact, to clear up anything you did not understand, and to keep your own record of meetings. The decision to submit is yours. Edinburgh makes the student fully responsible for the quality of the thesis, and says a supervisor's view that the work is ready "carries no guarantee" of the result (University of Edinburgh, 2025). Melbourne's hub is blunt about the same point: supervisors are there to support you, not to take charge of the research.
At the start, Glasgow asks you and your team to agree how often you will meet, the supervision you need and the time each side can give. Melbourne requires a written supervisory agreement within three months of starting (University of Melbourne, 2026).
How often should you meet your supervisor?
As often as your university's code says, and the codes disagree by a wide margin. The table sets out what each one requires on meetings, feedback and records.
| University | Meetings | Feedback on written work | Records |
|---|---|---|---|
| Loughborough (dissertations and final projects) | At least one hour one to one per 10 credits of the module | The handbook must say whether a full draft or one chapter is read, and when checking stops | The handbook must say how records are kept |
| Imperial (research degrees) | Normally at least one hour a week on average | Returned "in reasonable time" | No rule on meeting records in the 2016 code |
| Bristol (research degrees) | Formal meetings at least monthly, normally in person | "Promptly"; no full proofread | The main supervisor records the date and format of each formal meeting |
| Edinburgh (research degrees) | Contact at least twice in every three months | Discussed at supervision meetings | You send a note of each key meeting for agreement |
| Glasgow (research degrees) | Agreed between you and the team | "In reasonable time" | You keep one; concerns are recorded and agreed by both |
| Toronto (doctoral) | The committee meets you at least once a year | A lengthy chapter in under a week is unreasonable; over a month without notice is too | The committee writes a formal report |
| Melbourne (graduate research) | An agreed schedule of regular individual meetings | Within a reasonable agreed timeframe | A written supervisory agreement, reviewed during candidature |
| Stanford (graduate) | Each programme sets minimum advising expectations every year | Set by the programme | Set by the programme |
The sources for each row are Loughborough University (n.d.), Imperial College London (2016), University of Bristol (n.d.-b), University of Edinburgh (2025), University of Glasgow (2025), University of Toronto School of Graduate Studies (n.d.), University of Melbourne (2026) and Stanford University (2018).
The gap is real: a weekly hour on average at Imperial, contact twice a quarter at Edinburgh, though Edinburgh adds that there may be more meetings in the first months, while the project is being defined. Bristol's monthly minimum applies to part-time students too. If your code names no figure, as Glasgow's does not, agree one in writing at the first meeting.
How do you prepare for a supervision meeting?
Send something in advance, arrive with a short agenda, and write up what was agreed the same day. Edinburgh lets students call meetings, and Edinburgh and Glasgow both expect the student to produce the record, so the preparation is yours.
- Decide what the meeting is for: a decision, feedback on a draft, a problem, or a progress check.
- Send the draft or a one-page update early enough to be read; Glasgow asks for written work "in sufficient time" for comments.
- Before a draft goes, check its references in Verify references and its format against the handbook, so no comment is spent on a wrong year.
- List three to five questions, most important first, each one your supervisor can answer.
- Bring your plan and your deadlines, so dates can be agreed in the room.
- Afterwards, send a short note of what was decided and who does what by when.
What to send in advance
A draft with a covering note saying what reading you want: the argument, the structure, the methods, or one section. The Toronto guidelines expect supervisors to say whether they want finished sections only or will read outlines and partial drafts (University of Toronto School of Graduate Studies, n.d.). For a methods chapter, our methodology chapter guide shows what a reader looks for.
What the record should say
Edinburgh's list is short enough to copy: the date and purpose of the meeting, any problems identified, and the action points (University of Edinburgh, 2025). At Bristol the main supervisor logs the date of each formal meeting and whether it was in person; keep your own note of the decisions too. The formal records of a PhD, with its progress reviews and forms, are a subject we will cover separately.
How much feedback on drafts should you expect?
Enough to act on, within a time you have agreed, but not a line-by-line correction. The Toronto guidelines give the only figures in any code read here: a thorough reading of a long chapter in less than a week "would probably be unreasonable", and so would a month without a reply unless the supervisor gave notice. Where drafts have already been read and the comments acted on, the final version should be read within two or three weeks.
For undergraduate and taught projects, the allowance is set locally. Loughborough University (n.d.) makes each school's handbook state how much a supervisor will read, a full draft or a single chapter, and the point after which written work can no longer be checked; the dates and draft allowances at other universities are compared in the dissertation hub.
Asking for feedback you can use
Ask a question the draft can answer: "Does chapter three answer research question two?" gets a better reply than "any thoughts?". Say when the next draft is due, so the supervisor can plan their reading.
Using what comes back
Keep a list of every comment and what you did about it: changed, answered, or disagreed with and why. It makes the next meeting a check of progress, and it is your evidence if a disagreement goes further.
What if you disagree with your supervisor or want to change?
Raise it with your supervisor where you can; every code read here names who to see next. A disagreement about the work is normal and is settled in the meeting, with your reasons in the record. A breakdown of the relationship has a formal route.
The formal routes
- Imperial: the Director of Postgraduate Studies first, or the departmental Senior Tutor (PGR) (Imperial College London, 2024).
- Glasgow: a member of the supervisory team, then the Postgraduate Convener or Director, Head of School or Adviser of Studies, then the Dean of Graduate Studies, with signed and dated copies of the written records sent to the Graduate School Office (University of Glasgow, 2025).
- Edinburgh: the School Postgraduate Director, then the Secretary or Chair of the College Committee; the Students' Association's Advice Place also helps (University of Edinburgh, 2025).
- Melbourne: a member of your advisory committee first, for anything you cannot raise with your supervisors.
- Stanford: a resource each programme designates, such as the department chair or the Director of Graduate Studies, and campus offices including the University Ombuds (Stanford University, 2018).
Changing supervisor
Possible almost everywhere, on request. Stanford lets the student or the advisor ask for a change when research diverges or "irreconcilable differences" arise, and expects the programme to find another advisor, which may mean modifying the project. Melbourne's policy allows the candidate, a supervisor or the committee chair to start a change (University of Melbourne, 2026), and its hub asks you to discuss the issue with your supervisors first where possible. Glasgow warns that a replacement may be less expert in a narrow field, and that the project may have to adapt.
How do co-supervision, committees and PhD panels work?
Most research students have more than one supervisor; what differs is who leads.
| University | Structure |
|---|---|
| Edinburgh | At least two supervisors: a principal, and a co-supervisor or assistant supervisor |
| Glasgow | A primary supervisor with primary responsibility, and one or more secondary supervisors |
| Imperial | A main supervisor as the single point of contact, possibly with co-supervisors |
| Melbourne | At least two supervisors, plus an advisory committee of at least three |
| Michigan | A dissertation committee of at least four, with a chair and a cognate member from outside the home programme |
| Toronto | A supervisor and a supervisory committee that meets you at least once a year |
In co-supervision at Edinburgh, used for interdisciplinary work, the co-supervisor plays a major role and the split of duties is agreed at the outset. Michigan's cognate member gives a non-specialist's view (University of Michigan Rackham Graduate School, n.d.). The glossary defines a supervisory team and a dissertation committee. With several readers, send one draft to all and take conflicting advice to a joint meeting rather than choosing alone.
What changes when supervision is remote?
The rules stay the same; the arrangements need agreeing more carefully. Glasgow asks students who rarely meet face to face to agree how and how often they will communicate, especially across time zones, and asks supervisors to say how long replies and feedback may take. Notes and agreed actions are kept as usual.
Edinburgh allows meetings online. Bristol permits online meetings where necessary and agreed with the main supervisor, but students on a visa must attend the monthly meetings in person (University of Bristol, n.d.-b). Imperial's precepts ask supervisors on collaborative programmes to keep contact normally every two weeks. Toronto adds that a supervisor on research leave should still read drafts on time, or an acting supervisor should be found.
What may a supervisor not do?
A supervisor may not write your dissertation, and none of the codes read here allows anyone else to. Edinburgh requires the thesis to be your own work except where the declarations page says otherwise. Bristol says supervisors give feedback on the dissertation's development but are not required to proofread the whole of it before examination, that a third-party proofreader must not produce text or alter its intellectual substance, and that asking someone else, or an AI application, to complete assessed work is contract cheating (University of Bristol, n.d.-a).
Glasgow draws the line for proof-readers: spelling, punctuation, grammar and layout, but never adding, rewriting, restructuring or fact-checking content, translating it, or cutting it to a word limit. Melbourne's policy holds any help with thesis preparation to the Australian Standards for Editing Practice. Read together, the codes put supervisors on the same side of that line: they comment, and you write. The glossary's entry on third-party proofreading sets out the terms.
How EdCitation fits between supervision meetings
EdCitation is the best tool for the checks a supervisor should not have to make, because it checks each reference against the publisher's record instead of writing one, and it keeps "could not check" apart from "not found". We ran a three-entry list through Verify references on 26 September 2026: one real paper, a second real paper with its year changed from 2009 to 2012, and one reference we invented.
- Anne Lee's 2008 paper in Studies in Higher Education came back verified: "Title, year and first author match a published record."
- Mainhard and colleagues' paper in Higher Education, given as 2012, came back doubtful, shown on the page as "Check this": "A record with this title exists, but it was published in 2009, not 2012."
- The invented Harrow and Pellegrini paper came back not found: "No publisher's or registry's record matches this reference."
That is a comment your supervisor no longer has to write. For the whole draft, References from a file reads the reference list out of the uploaded file, looks up every entry, matches every in-text citation to the list, and keeps the paper so the next run shows what changed since the last one; a free account includes three checks a month, and Pro, $8 a month, removes the limit. Check your paper reads your handbook or brief into a checklist, with word limits, required headings, the citation style and the formatting rules, each shown with the sentence it came from, free.
A free account keeps the sources your supervisor recommends in a list you can tag by chapter, and the Team page invites your supervisor by email to see your reading list; they create a free account to open it. Dissertation references from proposal to submission follows that list to the final week. None of it replaces the meeting: EdCitation checks what you wrote, and your supervisor judges whether it holds up.
Quick questions
How often should I meet my dissertation supervisor?
As your code or handbook sets. Imperial asks for about an hour a week for research students, Bristol for a formal meeting at least monthly, Edinburgh for contact at least twice in three months, and Loughborough for an hour of one-to-one supervision per 10 credits of a dissertation.
How long should a supervisor take to return feedback on a chapter?
A time you agree in advance. Toronto's guidelines call less than a week for a long chapter unreasonable, and more than a month without notice unreasonable too; most other codes say only "reasonable time" or "promptly".
Can I change my dissertation supervisor?
Usually, yes. Stanford and Melbourne let the student ask for a change, and Glasgow, Edinburgh and Imperial set out who to see if a problem cannot be settled with the supervisor.
Will my supervisor proofread or edit my dissertation?
No. Bristol says supervisors are not required to proofread the whole dissertation, and the codes read here expect the work to be your own. Check the references yourself first, free, in Verify references.
Who keeps the record of supervision meetings?
Usually you. Edinburgh asks the student to send a note of each key meeting for agreement, and Glasgow expects students to keep a record; at Bristol the main supervisor logs the date and format of formal meetings.
References
- Imperial College London. (2016). Code of practice for research supervisors. https://www.imperial.ac.uk/media/imperial-college/administration-and-support-services/registry/academic-governance/public/academic-policy/codes-of-practice-for-research-degrees/Code-of-Practice-for-Research-Supervisors.pdf
- Imperial College London. (2024). Research degree precepts and guidance 2024-25. https://www.imperial.ac.uk/media/imperial-college/administration-and-support-services/registry/academic-governance/public/academic-policy/research-degree-examinations/Research_degree_precepts_and_guidance_2024-25.pdf
- Loughborough University. (n.d.). Code of practice governing supervision of dissertations/final individual student projects. Academic Registry. https://www.lboro.ac.uk/services/registry/aqphandbook/documentation/cop-supervision-student-diss/
- Stanford University. (2018). 3.3.1 Academic advising: Policy. Graduate Academic Policies and Procedures. https://gap.stanford.edu/handbooks/gap-handbook/chapter-3/subchapter-3/page-3-3-1
- University of Bristol. (n.d.-a). Academic integrity and research degrees. Academic Quality and Policy Office. https://www.bristol.ac.uk/academic-quality/pg/code-of-practice/assessment/academic-integrity/
- University of Bristol. (n.d.-b). Supervisory support and engagement. Academic Quality and Policy Office. https://www.bristol.ac.uk/academic-quality/pg/code-of-practice/supervision/support/
- University of Edinburgh. (2025). Code of practice for supervisors and research students. Registry Services. https://registryservices.ed.ac.uk/sites/default/files/2025-07/Supervisors%20and%20Research%20Students,%20Code%20of%20Practice%20-%20print-friendly%20version_0.pdf
- University of Glasgow. (2025). PGR code of practice 2025-26. https://www.gla.ac.uk/media/Media_1217517_smxx.pdf
- University of Melbourne. (n.d.). Working with my supervisors. Graduate Research Hub. https://gradresearch.unimelb.edu.au/roles-and-responsibilities/working-with-my-supervisors
- University of Melbourne. (2026). Graduate research training policy (MPF1321). https://policy.unimelb.edu.au/MPF1321/
- University of Michigan Rackham Graduate School. (n.d.). Guidelines for dissertation committee service. https://rackham.umich.edu/faculty-and-staff/dissertation-committees/guidelines-for-dissertation-committee-service/
- University of Toronto School of Graduate Studies. (n.d.). Supervision guidelines for faculty, section 4: Effective supervision in practice, from the initial stage to finishing up. https://www.sgs.utoronto.ca/supports/supervision/supervision-guidelines/supervision-guidelines-for-faculty-section-4-effective-supervision-in-practice-from-the-initial-stage-to-finishing-up/