Research ethics in supervision

About the author: Line Wittek is a Professor at the University of Oslo’s Department of Education. 

Research supervision is a complex and challenging process, with many potential pitfalls along the way. Research ethics guidelines provide valuable tools for reflecting on and improving our own supervisory practices. However, their broad and rather general formulation means they require contextual interpretation to be fully effective. 
 
This article explores the role of research ethics guidelines in higher education supervision, focusing on three key areas: (a) roles and responsibilities, (b) relational dynamics between researchers and students, and (c) the collective responsibility of the research community. Each section begins with a case, which is analysed and explored through the lens of ethical principles and current research on supervision. While the cases are drawn from PhD supervision, many of the primary insights also apply at MA and BA level, particularly in the discussions in section (b) on relational dynamics and section (c) on collective responsibilities. 
 
The cases are based on real-life scenarios from the university and college sectors, though some details have been modified to protect anonymity. 

(A) Clarifying roles and responsibilities 

The first case brings together a number of research ethics issues, particularly those relating to the lack of clear expectations on one hand and ineffective communication about roles and responsibilities on the other. 

Case A 

Anne, is employed as a research fellow on a large interdisciplinary EU project. The project leader agrees to act as her primary supervisor, and a co-supervisor from another department at the same university is also appointed. However, the co-supervisor’s role is not clearly defined, and no explicit up-front agreements are drawn up determining who has access to and the right to use the data Anne collects, or who will be credited as co-authors for her publications. 
 
Communication between Anne and the project leader is poor, and after a few months a new primary supervisor is appointed. Unfortunately, the new supervisor also fails to address key questions about data access and co-authorship.  
 
Anne’s first two articles are published successfully. She includes the project leader and both supervisors as co-authors, with herself listed first and the project leader last. When her third article is ready for publication, Anne shares the draft with the project team. The project leader (her former supervisor) is highly critical of the way the research has been carried out, claiming the research analyses are both unreliable and invalid. He wants the data to be reanalysed and refuses to submit the article for publication in its current form. He also requests access to the data. Anne counters, questioning whether the project leader even qualifies as a co-author given he has not made any significant contributions to the project. She is unwilling to redo the analyses, and her current supervisors believe the article meets the necessary standards for reliability and validity. 
 
The authorship order becomes another bone of contention. The new primary supervisor opts out of being listed as a co-author, feeling he has not contributed significantly to the publications. Meanwhile, the co-supervisor argues that she should be listed last, suggesting her name would enhance the reliability and credibly of the research and promote gender balance. 

A persistent issue in this case is the lack of clarity around roles and responsibilities. While Anne is ultimately accountable for her research, her supervisors, the project leader and other team members could have avoided this unfortunate situation by addressing key questions at an early stage of the process. But what exactly are the responsibilities of a project leader, primary supervisor, co-supervisor and research fellow? In interdisciplinary projects like this, where it is not uncommon for participants to come from diverse research traditions, it is particularly important to establish clear roles and expectations. Established researchers may naturally revert to “just doing things like we always done”, but for newcomers or participants from other institutions, such practices can be shrouded in mystery. Two published articles indicate that Anne is working well on the project, making good progress and receiving adequate supervision. However, the project leader, having delegated supervision to two colleagues, now disputes their assessment of fundamental aspects of the research just as the work is nearing completion. Anne finds herself caught between highly conflicting feedback. While her supervisors have given the green light for publication, her project leader (and her former primary supervisor) insists the entire analysis needs to be redone. As a newcomer to the research community, how should Anne interpret the conflicting feedback she has received? How should she weigh the expertise of different senior researchers against each other and make decisions in the midst of this conflict 
 
Interpreting feedback can be inherently difficult when the provider and recipient come from vastly different backgrounds. In the context of academic supervision, this has been likened to “learning a new language” (Ajjawi and Boud, 2018, p. 1106). The contradictory nature of the feedback Anne has received puts her in an exceptionally tough position.  
 
Another issue highlighted in this case is the question of access to and the right to use data. The senior researchers have not flagged use of data as a discussion point at any time during the project, but as Anne’s PhD project nears completion, serious problems emerge. The project leader has demanded that Anne redo her analyses, requested access to her data and argued that submission of her third article should be delayed. Is this a reasonable request, and who truly owns the rights to Anne’s data? 
 
Ethical guidelines place a particular responsibility on research supervisors to protect the rights of co-authors when students, PhD candidates and junior researchers are involved, due to the asymmetric power relations (NESH 2021). Early clarification is particularly important in large, interdisciplinary projects with multiple authors and in international collaborations, where differing practices and expectations are more likely to arise (ALLEA 2023, Vancouver Recommendations 2023, sections 16, 17 and 18 of the Montreal Statement 2013). 
 
A third point of contention is deciding who should be listed as a co-author and the authorship order. What criteria should guide these decisions? Co-authorship agreements and authorship order agreements should not be left solely to the research fellow and supervisors; they should be determined collectively by the research team (NESH 2021, Section 9). If roles and responsibilities change during the project, these agreements need to be revisited to ensure fair recognition and involvement. Everyone who has significantly contributed to the project should have the opportunity to participate in the final stages leading to publication (Vancouver Recommendations 2023). However, simply contributing to data collection, supervision or funding does not meet the threshold for co-authorship. In Anne’s case, the co-supervisor argues for inclusion as a co-author on the third article, not because of a significant contribution to the work but because her name would add some prestige to the publication and support gender balance. This reasoning falls outside established co-authorship guidelines. Research Integrity in Norway’s 2019 study revealed that many PhD candidates reported being pressured into “gift authorships”, i.e. listing senior researchers as authors despite the fact that they had not significantly contributed to the content. 
 
“Researchers shall respect the contributions of others and comply with recognised norms of co-authorship and collaboration.” (NESH 2021, Section 9). In this case, the project leader’s approach to collaboration can hardly be described as respectful towards either the candidate or the other researchers. Project leaders have extensive expertise and carry significant responsibility and high expectations for the project’s outcomes. In addition, project leaders’ obligations to funders and collaboration partners can place both intentional and unintentional pressure on fellow researchers to meet specific project goals. Such demands and expectations can thus compromise individual researchers’ independence. When students or PhD candidates are involved, particular attention should be paid to the fact that power imbalances can exacerbate the risk of undue influence or conflicts of interest, increasing the likelihood that research outcomes are skewed. 
 
A Finnish interview study of candidates and their supervisors across various disciplines (Löfström and Pyhältö, 2017) found that many candidates felt pressured to adopt their supervisor’s perspectives or interpretations. Supervisors, meanwhile, often expressed difficulty in striking the “right” balance between supporting the candidate’s wishes, interpretations and decisions, and offering advice informed by their own expertise and research experience. Finding this balance is no small task. As supervisors, our job is to guide candidates through the many complex decisions they face during their PhD projects. This also involves giving specific recommendations and advice, such as on refining research questions or identifying relevant source materials. At the same time, we must safeguard the researcher’s (here the research fellow’s) actual independence (NESH 2021). However, this responsibility does not fall solely on the supervisor and candidate; everyone involved in a project has a duty to protect researchers’ independence (Section 2).  
 
To ensure early clarity on these matters, researchers must be well-versed in the relevant research ethics guidelines and fully understand how to apply them within their specific field. Project leaders bear particular responsibility for ensuring that these discussions receive attention and time is allocated for them for all people involved in the project. However, it is equally critical for those with secure positions within the organisation to support those in more precarious roles, ensuring they are not unfairly pressured. Agreeing a set of “research ethics ground rules” can provide valuable guidance for navigating difficult decisions and help prevent conflicts like the one described in the case above. Such agreements should in particular safeguard the interests of junior researchers (ALLEA 2023, Section 2.1). Supervisors should be particularly attentive to asymmetrical power relations, and not use their academic authority to their own advantage or in ways that violates other people’s integrity (NESH 2021, Section 5). 

(B) Relational dynamics of supervision 

Case B highlights the relational dynamics in supervision and the impact of trust, or the lack thereof. It also raises questions about a supervisor’s responsibility to create a culture of ethical research practices.  

Case B 

Martin is a research fellow working on a project led by Marie. Marie’s research team has secured funding for several large-scale projects. Already supervising five research fellows, she takes on Martin as well, feeling that this is a natural extension of her role as project leader. Pelle, a member of Marie’s research team, has served as a co-supervisor for several of Marie’s research fellows in the past. During a lunch break, Pelle suggests to Marie that they arrange a joint supervision meeting with Martin to clarify their respective roles and expectations. Marie does not respond, but checks her watch and says she has to rush off to another meeting. 
 
One day, Martin approaches Pelle in his office, asking for a confidential conversation. He starts by complaining about Marie’s lack of availability, explaining that she only gives superficial feedback on the drafts he submits. When he broached the subject with her, she became irritated stating that supervisors shouldn’t be expected to “mother” their research fellows, like many young people seemed to believe.  
 
Martin is also concerned about his rights to the empirical data he is collecting. He has spoken with other research fellows on Marie’s projects, and several have said she used their data in her own publications without consulting them. Pelle, based on his own experience as a co-supervisor for Marie’s research fellows, has suspected for some time that her supervision leaves something to be desired. Too intimidated to bring up his concerns with Marie again, Martin concludes by saying, ”Pelle, you are my co-supervisor and you know Marie as both a colleague and a professional. Could you talk to her for me?”   

Marie is a highly accomplished researcher who accordingly supervises multiple research fellows. It is understandable that she needs to manage her time carefully and prioritise rigorously. However, no matter how tight her schedule, there are certain responsibilities she cannot neglect as a supervisor. One of her key roles is to ensure that her candidates learn to conduct research in line with the core principles of research ethics as recognised by the international academic community: reliability, honesty, respect, and accountability (ALLEA 2023). This includes a responsibility to highlight and help the candidate learn both the content of the guidelines and how to put the guidelines into practice – or quite simply, to  be a good role model. 
 
Relational dynamics are another component thematised in the guidelines. Supervisors are expected to be attuned to the vulnerabilities that early-career researchers can often experience (NESH 2021, Section 2). However, this responsibility should not be interpreted as resting solely on the supervisor. The research community as a whole has a duty to establish a working environment characterised by mutual respect, equality, diversity and inclusion. Supervisors, in particular, must remain conscious of the power they wield: “Supervisors and students/PhD candidates must treat each other with respect. Supervisors should not misuse their position to their own advantage, (NESH, Section 5), This applies to both academic and personal matters.” (NESH, Section 5). In the case described, Martin feels neither respected nor included by his supervisor. Far from receiving “care”, he feels rejected and is afraid to reach out to his primary supervisor again. But what actually makes a supervisory relationship effective? And what is key to establishing it and making it an effective resource for the candidate’s development? Below, I explore some key insights from research in this area. 
 
Several research studies show a correlation between the frequency and length of supervision meetings and the overall quality of the supervision (Lee 2018; Ajjawi and Boud 2017; Löfström and Pyhältö 2017). Candidates who benefit from well-functioning supervisory relationships learn more, grow professionally and develop stronger generic skills compared to their peers. Supervision is also the most significant factor influencing the time candidates take to complete their work and the quality of the final outcome. 
 
A candidate’s sense of satisfaction during their research period and their feeling of success as a research fellow are closely tied to their relationship with their supervisor and the trust that develops within it (Wittek 2023; Ajjawi and Boud 2017). Providing feedback on drafts, as Marie does, is a core aspect of research supervision. However, here Martin finds Marie’s feedback unhelpful and superficial. In recent years, the focus of supervision research has shifted significantly (Wittek 2023; Ajjawi and Boud 2018). While earlier research concentrated on how supervisors phrase and communicate feedback, increasingly the emphasis has shifted towards the relational and contextual aspects of feedback. The question has become less about how feedback is given and more about how it is received, interpreted and applied by the recipient. Feedback only becomes valuable when the recipient can directly connect it to their specific needs in their work (Ajjawi and Boud 2017; Telio et al., 2015). This shift emphasises the importance of the candidate or the student taking an active role in the supervisory relationship. Candidates need to ask questions, articulate their struggles and actively consider their interpretation of the supervisor’s input. At the same time, candidates and students must obviously be open to constructive feedback and willing to use it to improve their work, approach or thinking. But what really determines whether a candidate actively engages with feedback in this way? 
 
Telio et al.’s 2016 study offers valuable insights in this context. The researchers investigated what helps medical students in the final stages of their training to embrace and trust feedback from their clinical supervisors. The students’ perception of their supervisor’s credibility proved to be a particularly important factor (Telio et al. 2016, p. 937–938). It might be natural to assume that the perception of credibility would be most closely tied to the supervisor’s academic expertise. However, three other quite different factors stood out. Firstly, students highlighted the importance of the supervisor’s genuine commitment to the feedback process. Secondly, they valued whether the person giving the feedback showed a personal interest in them. Thirdly, it was crucial that the supervisor came across as friendly, approachable, making the students feel respected and liked as individuals. 
 
Molloy and Bearman (2018) argue that the development of trust in the supervisory relationship is closely linked to the complex dynamic between being vulnerable on one hand and appearing professional and competent on the other. Here, both supervisor and student intuitively seek a meaningful balance where they can safely and frankly admit struggles and ask for help without risking losing face. According to Molloy and Bearman (2018), vulnerability is essential for learning, precisely because it is only by making ourselves vulnerable that can we challenge ourselves and change our views: 
 
We have to open ourselves up to the possibility of imaging things differently; this is more than the incremental learning of adding new facts to old ones. (…). Rather, this is learning as transformation: different ways of knowing or being; profound epistemic or ontic changes in how we make sense of the world (p. 3). 
 
Martin’s situation also raises challenges relating to inter-collegiate relationships. Pelle is a co-supervisor for a research fellow who feels there are major shortcomings in the supervision he is receiving from his primary supervisor, and asks Pelle to intervene on his behalf. Pelle has long suspected that Marie’s research fellows are not receiving the guidance they need, and that she rarely integrates research ethics principles into her supervision. Now, those suspicions have been confirmed. What should Pelle do in his role as co-supervisor? Should he speak with Marie? Should he actually have stepped in much earlier, given his long-held suspicions? Should he involve others in the organisation, and if so, whom? In the following section, we will have a closer look at the research community’s collective responsibility in providing effective supervision. 

(C) The collective responsibility of the research community 

Supervision has traditionally been viewed as an individual responsibility (de Lange and Wittek 2023). However, recent studies show that many supervisors, whether at PhD or MA level, feel uncertain about their role and isolated in managing its demands (Wittek and de Lange 2021; Lee 2018). When challenges arise in supervisory relationships, supervisors often find themselves navigating them alone, as demonstrated in Case C:  

Case C 

Kari, an associate professor at a small university, was assigned as the primary supervisor for a research fellow a year ago. The research fellow lives far away from the university and only visits occasionally. So far, Kari has only spoken to the research fellow twice, and both conversations were relatively brief, with the candidate coming across as quite dismissive on both occasions. Kari has sent several emails encouraging the candidate to get in touch. However, the candidate has merely explained that she does not have many questions and that if she does normally consults her mother, a professor in the same field at a major university in her hometown. Kari feels unsure about her role as a supervisor. Supervision challenges are rarely discussed in her academic environment. ”It’s just assumed you’ll figure it out on your own,” she says. Kari is unsure how to proceed. She feels rejected by the candidate, and is beginning to doubt her own abilities. She is also unclear on the formal guidelines for PhD supervision. She has sought advice from colleagues, but their answers were evasive. 

The ALLEA guidelines emphasise that the research community has an ongoing responsibility to provide research supervisors with proper training and information about research ethics (ALLEA 2023, p. 6). Furthermore, researchers shall contribute to building academic communities characterised by openness, factuality and collegiality. (NESH 2021, p. 11). In Kari’s case, it is evident her institution is falling well short in this area. 
 
The responsibility for ensuring that participants learn about and are equipped to uphold research ethics principles in practice is shared across different levels of the organisation. Institutions are tasked with providing training on research ethics for all supervisors. As supervisors, we are responsible for staying up to date with current guidelines and any changes. As members of the research community, we also have a collective responsibility to operationalise these guidelines, monitor practices and speak up if we notice anything that could potentially lead to a breach of ethical research guidelines further down the line (ALLEA 2023, p. 4). This also applies to situations or relationships where we are not directly involved.  
 
PhD candidates are accountable for their own research and shall enjoy “individual freedom and have real independence” (NESH 2021, p. 10). But does this mean a candidate can forgo supervision altogether? What about the supervisor’s duty to guide candidates in conducting ethical research? How can Kari take responsibility for a project she knows little about? And how can she support the candidate as a newcomer to academia? In the case above, there is a clear lack of explicit guidelines defining the roles and responsibilities of both the candidate and the supervisor. Both the research fellow and the supervisor share a collegial responsibility as members of the academic community. “Lecturers, supervisors, and project leaders have a particular responsibility for including students and PhD candidates in the scientific community and for introducing them to research ethics.” (NESH 2021, Section 2). This requires both parties to work together to clarify their roles and commitments within the supervisory relationship. At the same time, it is crucial to acknowledge the inherent power relationship. Additionally, it must be emphasised that the institution has a key responsibility to establish clear structures that ensure research supervision is conducted in accordance with recognised research ethics standards. In practice, this may involve developing procedures, written guidelines or requirements that are regularly reviewed, discussed and revised. 
 
A core principle of research ethics is that research must be conducted with openness and independence. In the context of supervision, this requires that the supervisor and candidate communicate with each other, and that their communication is built on trust and openness, and that this creates room for academic disagreement, constructive critique and ethical deliberation both around research and ethical principles in research (NESH 2021, p. 11). Explicit agreements about roles, ownership and responsibilities must be established from the outset. As the guidelines state, “if responsible agreements cannot be reached or maintained, the research should not proceed” (p. 34). In other words, while having checklists and well-documented agreements is vital, they are not sufficient on their own. Proper agreements matter, but they cannot replace the relational foundations of trust, attentiveness and mutual respect. 

Summary 

For ethical principles to fully inform supervision practices in research, they must be made relevant and actionable across individuals, disciplines and academic environments: “An effective European Code of Conduct for the research community promotes an ethical mindset. Its principles are relevant across the research system and in all disciplines” (ALLEA 2023, p. 4). Organisations that embrace an ethical mindset ensure that all supervisors are familiar with research ethics laws and guidelines and cultivate a deep collective understanding of how these principles can be applied in supervision practices. Responsibility for ethical supervision must be shared across all levels of the organisation, and in research supervision between research fellows, supervisors, project leaders and institutional leadership alike. 
 
Key topics that should remain central to ongoing discussions about ethics in supervision include: 

• Clarifying roles, ownership and data rights 
• Clarifying co-authorship 
• Building trust in the supervisor-candidate relationship  
• Viewing ethics in supervision as a collective responsibility 

Translating ethical guidelines into practical action requires contextual interpretation by the research community as a whole. Such interpretations should be shaped within each research group, rather than as a one-off decision, but as part of an evolving and continuous conversation. In practice, this means revisiting the mentioned topics whenever new projects are launched with new participants and whenever PhD candidates, junior researchers or students are brought into the fold. 
 
At the heart of it all, the most important thing is to approach each other with honesty, respect and trust. This requires more than just knowledge of standards and guidelines. As members of the research community, we must be willing to “put ourselves out there” and be willing to listen and consider others’ perspectives. These qualities are difficult, if not impossible, to instil through research ethics courses or to capture in policy documents or formal strategies. Instead, we must all take responsibility for continuously addressing this area within our respective academic cultures. 

References 

Ajjawi, R. & Boud, D. (2017). Researching feedback dialogue: an interactional analysis approach. Assessment and Evaluation in Higher education, 42 (2), 252–265. doi: 10. 1080/02602938.2015.1102863 
 
Ajjawi, R. & Boud, D. (2018). Examining the nature and effects of feedback-dialogue. Assessment & Evaluation in Higher Education, 43 (7), 1106–1119. doi: 10.1080/02602938.2018.1434128 
 
ALLEA (2023). The European Code of Conduct for Research Integrity – Revised Edition 2023. Berlin. DOI 10.26356/ECOC 
 
Montreal Statement on Research Integrity in Cross-Boundary Research Collaborations 
https://www.wcrif.org/guidance/montreal-statement 
 
De Lange, T, Wittek, L. (editors), (2023). Faculty Peer Group mentoring i Higher Education: Developing Collegiality through Organised Supportive Collaboration. Springer. 
 
Firing, K., Klomste, A.T. and Moen F. (2013). Masterstudenters opplevelse av møter med veileder: Det er veiledningen som gjør at en føler at en mestrer. UNIPED. 36(2), 81–92. https://doi.org/10.3402/uniped.v36i2.21515
 
Lee, A. (2018). How are doctoral students supervised? Concepts of doctoral research supervision. Concepts of doctoral research supervision, Studies in Higher education, 33 (3), 267–281, DOI: 10.1080/03075070802049202 
 
Löfström, E. and Pyhältö, K. (2017). Ethics in the supervisory relationship: supervisiors` and doctoral students dilemmas in the natural and behavioral sciences. Studies in Higher education, 42 (2), 232–247 
 
Mahmud, S., and T. Bretag. (2013). Postgraduate Research Students and Academic Integrity: «It`s about Good reserch training». Journal of Higher education Policy and management 35 (4), 432–443 
 
Molloy, E. & Bearman, M. (2018). Embracing the tension between vulnerability and credibility: “intellectual candour” in health professions education. Medical education, S.1–10. doi: 10.1111/medu.13649 
 
NESH (2021). Guidelines for Research Ethics in the Social Sciences and the Humanities. ISBN: 978-82-7682-103-1 
https://www.forskningsetikk.no/en/guidelines/social-sciences-and-humanities/guidelines-for-research-ethics-in-the-social-sciences-and-the-humanities/ 
 
RINO, tredje delrapport: Forskningsintegritet i kontekst: Resultater fra en kvalitativ studie (2019). Drivdal, L., Kaiser, M., Hjellbrekke. J., Rekdal, O.B., Skramstad, H., Torp, I.S., Ingierd, H. ISBN trykt utgave: 978-82-7682-095-9 
https://www.forskningsetikk.no/ressurser/publikasjoner/forskningsintegritet-i-kontekst-resultater-fra-en-kvalitativ-studie/ 
 
Telio, S., Ajjawi, R. & Regehr, G. (2015). The «Educational Alliance» as a Framework for Reconceptualizing Feedback in medical Education. Academic medicine, 90 (5), May. S. 609-614 
 
Telio, S., Regehr, G. & Ajjawi, R. (2016). Feedback and the educational alliance: examining credibility judgements and their consequences. Medical Education, 50, 933–942. doi: 10.111/medu.13063. 
 
Vancouver recommendations (2023): https://www.icmje.org/icmje-recommendations.pdf 
 
Wittek, L. (2023). Feedback in the context of peer group mentoring: A theoretical perspective. I: De Lange, T. & Wittek, L. (editors), (2023). Faculty Peer Group mentoring i Higher Education: Developing Collegiality through Organised Supportive Collaboration. Springer. In press. Kapittel 2. 
 
Wittek, A.L. & de Lange, T. (2021). Action learning as means for supervisor development. I Lee, A. & Bongaardt, R. (eds.), The Future of Doctoral Research. Challenges and Opportunities. Routledge. ISSN 9780367858490. S. 249-262.  

Thanks to Bjørn Torgrim Ramberg and Vidar Enebakk for important input during the work on this text.