Research Project Assessments and Supervisor Marking

By Anosh Butt

Updated Apr 03, 2026

Introduction to Problem

Who should mark a dissertation: the supervisor who knows the project best, or an assessor who can stand further back? This Newcastle University study explores that tension and suggests that robust reconciliation, rather than simply removing supervisors from the process, is what protects academic rigour.

The paper examines undergraduate report marking where two assessors are involved, but where institutions do not always use the same model (McQuade et al., 2020). Some use two independent assessors and exclude the project supervisor because critics worry that familiarity may create leniency, a concern that also appears in work on transparency in undergraduate dissertation supervision. Others retain the supervisor because subject expertise can be vital when judging scientific argument and disciplinary depth. At Newcastle University, undergraduate biosciences programmes used the supervisor as one of the assessors, while postgraduate programmes did not. This gave the authors a useful way to compare supervisor involvement, project marks from 2011/12 to 2016/17, and the effect of two different reconciliation processes.

Solution

The authors analysed project marking data from undergraduate and postgraduate biosciences programmes delivered between 2011/12 and 2016/17. The undergraduate dataset covered 1,460 projects, with 897 from the four academic years 2011/12 to 2014/15 under reconciliation method I, and 563 from the two academic years 2015/16 to 2016/17 under reconciliation method II. The postgraduate dataset covered 894 project reports from 2011/12 to 2015/16.

Undergraduate project reports were presented in scientific paper format with a 5,000-word limit, while postgraduate reports were presented in thesis format with an 8,000-word limit. The pass mark was 40% for undergraduate projects and 50% for postgraduate projects. Across both levels, up to 150 academics from five research institutes and three schools were involved as supervisors and markers (McQuade et al., 2020).

The first reconciliation method required the two markers to submit independent marks, each supported by written comments, before discussing the work. It was used for undergraduate report reconciliations between 2011/12 and 2014/15, and for postgraduate reports between 2011/12 and 2015/16. The second reconciliation method also involved discussion, but the two markers had to return a single agreed mark with one set of supporting comments. It was used for undergraduate report reconciliations between 2015/16 and 2016/17.

The researchers divided the data into three groups: undergraduate project report marking from 2011/12 to 2014/15, postgraduate project report marking from 2011/12 to 2015/16, and undergraduate project report marking from 2015/16 to 2016/17. They then analysed the difference between the grades awarded by marker one and marker two. For undergraduate reports, marker one was the supervisor and marker two was another academic. For postgraduate reports, marker one was an internal examiner with relevant subject knowledge and marker two was an external examiner with appropriate expertise. For institutions, that design is useful because it helps separate the effect of supervisor involvement from the effect of subject expertise and reconciliation practice, including in discussions of how biology assessments can be fair and consistent.

Measurable impact

Across four years of undergraduate marking data, the mean difference between supervisors and faculty academics was +2.3% (McQuade et al., 2020). That suggests broad agreement overall, but with a modest tendency for supervisors to award slightly higher marks than second markers. There were still cases where differences between markers reached 30%, which shows that the more important issue is not simply supervisor bias, but the presence of disagreements large enough to require robust reconciliation.

The postgraduate results add an important qualification. Even when first markers came from the same internal pool of academics but did not know the students, discrepancies still reached 30%. That suggests familiarity is not the only source of variation. The level of subject knowledge, marking experience, and the elements of the work that each marker prioritises can all shape the final mark. It is also relevant that undergraduate second markers and postgraduate external assessors each marked more dissertations than supervisors or internal markers, often across a wider range of topics, which may have affected subject familiarity and marking reliability.

The authors report evidence that subject knowledge may be less essential when markers are assessing general features such as document structure, presentation, writing style, and referencing, rather than scientific content. They also highlight that assessors without relevant subject expertise appeared more cautious and were less likely to use the full marking range. On average, they marked significantly lower than assessors with subject-specific knowledge.

McQuade et al. (2020) also note that reconciliation comments referred to originality, breadth and depth of scientific knowledge, independent critical thinking, and scientific argument. That matters because higher education has faced pressure to move away from using undergraduate supervisors in dissertation marking, largely because of perceived leniency linked to contact and familiarity. The practical takeaway is that institutions should strengthen reconciliation and moderation processes wherever marks diverge, rather than assume that removing supervisors alone will make marking fairer.

FAQ

Q: How do the reconciliation methods affect the final grades of the projects?

A: The reconciliation methods matter because they determine how marker disagreement is resolved. Method I preserved two independent marks and written comments before discussion, while method II required a single agreed mark after discussion. The paper does not present a simple before-and-after result showing that one method always produces better final grades, but it does show that a clear reconciliation process is essential when differences between markers are large.

Q: What are the implications of these findings for future marking policies at Newcastle University and potentially other institutions?

A: The findings suggest that institutions should be cautious about assuming that excluding supervisors automatically creates fairer marking. The stronger policy lesson is to use robust reconciliation, clear assessment criteria, and effective moderation when marks differ widely, alongside work on staff-student partnerships to enhance assessment literacy. For Newcastle University and others, the study offers a way to review where subject expertise is genuinely needed and where process design is doing most of the work in protecting fairness.

Q: Was there any qualitative feedback from students or markers regarding the marking process and outcomes?

A: Not in the way the question implies. The paper analyses marking data and reconciliation comments, rather than collecting separate interview or survey feedback from students or markers. That means it offers evidence about how assessors behaved within the marking process, but much less evidence about how students or staff themselves experienced that process, a gap that also matters in the wider literature on student voice in assessment and feedback.

References

[Source Paper] McQuade, R., Kometa, S., Brown, J., Bevitt, D. and Hall, J., 2020. Research project assessments and supervisor marking: maintaining academic rigour through robust reconciliation processes. Assessment & Evaluation in Higher Education, 45(8), pp.1181-1191.
DOI: 10.1080/02602938.2020.1726284

Request a walkthrough

Book a free Student Voice Analytics demo

See all-comment coverage, sector benchmarks, and reporting designed for OfS quality and NSS requirements.

All-comment coverage with HE-tuned taxonomy and sentiment.
Versioned outputs with TEF-ready reporting.
Benchmarks and BI-ready exports for boards and Senate.

Prefer email? info@studentvoice.ai

UK-hosted · No public LLM APIs · Same-day turnaround