By Dr Sam Oakley, Researcher Development and Integrity Specialist

Both research and policy make it clear that supervisors have a key role to play for Research Integrity, but what does that actually mean in practice? ‘Research Integrity’ is a conceptual policy term that lacks a single definition: whilst acknowledging the different priorities of different disciplines, it is commonly used as a large umbrella concept to cover issues such as authorship, conflict of interest, plagiarism, and questionable research practices. It implies a passion for both high quality research and a positive culture for producing that.

As someone responsible for delivering training in ‘Research Integrity’ to new PGRs, it’s also my job to try and unpack some of the issues for supervisors so I can serve up something useful and practical for them. We need to be sure that the Research Integrity training for PGRs aligns with the messages we give to our supervisors.

At Glasgow, we view Research Integrity training as a starting point: a general introduction to the concept, our policies and the issues; the baton then has to be passed to our PGRs, and those who supervise them, to develop this understanding in a more subject-specific context.

We expect supervisors to take up this baton, so we need to make sure they are both expecting to do this work and supported and informed to be able to do it confidently. For this reason, supervisors are required to complete our Staff Research Integrity training course which includes a short module specifically on supervision. Then we offer an additional optional workshop to explore this further.

This blog post captures my current thinking on what supervisors might need and shares it for very welcome discussion.

Supervisor as Integrity guardian?

We make it clear that PGRs are responsible for the integrity of their own work, but supervisors have a responsibility to train them in good research practice. In many disciplines supervisors may well be co-authors on papers with their PGRs so it’s firmly in their interests to ensure that there are no issues with the quality of the work. I like to draw positive steps from a useful paper that explored the causes of misconduct in cases involving PGRs. This gives us the following tips:

• Be explicit about standards and expectations, ensuring that PGRs get support and training to meet these expectations.
• Review ‘source data’ (whatever that may look like in the discipline): not as a ‘spot check’ from the data police, but a positive and supportive check-in to reassure both of you that key concepts and processes are understood and being followed correctly.
• Minimise stress: “the prevalence for self-reported stress felt by over 50% of the accused PGRs”. Once again, we have the incontrovertible message that PGR wellbeing is key to the quality and success of their work!

Supervisor as villain?

When Sweden opened its research misconduct agency last year they reported that within their influx of cases, many were disputes between PhD students and their supervisors. There are tensions in this relationship that can take a wrong turn: expectations not met about publications, confusion over authorship decisions or who ‘owns’ ideas or work, for example. Many of these can be mitigated by clarity early in the process: conversations that pick through some of these areas of potential conflict. These are most useful (and fair) in the context of the wider research culture for the discipline.

Supervisor as victim?

The responsibility of the supervisor for their PGR has its limits but where these limits are drawn varies greatly (at one extreme, supervisors may get sacked if their PGR commits misconduct). However, there is inevitably a reputational risk (or worse, blame) if an issue does arise with the work of a PGR: how much did the supervisor know, why didn’t they spot a problem? This can be immensely stressful. Fortunately, actions to mitigate this risk align with good supervisory practice: keeping a record for both PGR and supervisor of:

• Reviews and discussion of ‘source data’ or processes.
• Discussions about expectations and standards.
• Explicit and active support for training and research governance (e.g. ethics, data management plans).
• Involving the PGR with wider team or group conversations, ensuring that they work within a positive research culture and have other sources of support.

Supervisor as Role Model?

“Role model of integrity”? As I type it, I can feel the weight of responsibility in that phrase and an awareness that it is yet another responsibility laid at the door of the supervisor. However, I’d like to offer my personal interpretation of this: of course, we expect all researchers to model good research practice, but I’d suggest these are the key actions for a supervisor to ‘role model’:

• A constant questing to interrogate current practice and find improvements: open research and reproducibility are obvious recent solutions for better transparency, for example.
• Keeping an open mind and humility. Never assuming to know all the answers, but being willing to question rigour, standards and ethics.
• Resistance to the inexorable pull to publish faster at the expense of quality, and work ever-harder at the expense of wellbeing.
• Actively promote kindness. A supportive and inclusive culture is essential for quality research, not just ‘nice to have’. It supports good wellbeing (and so good decision making) and enables the enrichment of ideas and perspectives that diversity brings.

It can be easy to forget to make your thought and decision-making processes explicit but modelling this for PGRs and (even better!) involving them in decision-making, is the best possible way that supervisors can support research integrity.

What does this look like for your discipline?

This is the question we pose in all our Research Integrity training. For my own PhD (Egyptology), if I had a time machine, I’d ask for conversations around ‘data management’ when the ‘data’ is ancient texts; the role of subjectivity in interpretation and how to steer clear of ‘cherry-picking’; how we communicate robust research to a public watching ‘Stargate’; how we acknowledge a hefty colonial past and address the issues ongoing with that. Exploring all those topics, ideally within a wider group, would have improved my scholarship tremendously.

So, over to you: what conversations do you need to start?