Nichola Burton, Program Manager (Institutional Underpinnings), ARDC
The Australian Research Data Commons (ARDC) works with Australian universities and other research institutions to support and uplift research data management capability. These activities include the recent Institutional Underpinnings program, which has brought together universities to improve national coordination around research data management. In this program, we investigated the ways that a university’s functional units and processes contribute to effective research data management. In those discussions, universities recognised research ethics committees as an important contributor to good research data management by researchers. Here I reflect on some of the issues and difficulties around research data management in the ethics review process.
Researchers have a responsibility to manage their research data responsibly, in a way that minimises the chances of any data loss, accidental release, theft, or mishandling that could have a negative impact on individuals, communities, ecosystems or organisations. Their data forms an important record of their research and underpins research integrity, and there are both regulatory and institutional requirements to store those records appropriately. Research data is also a valuable publicly-funded asset, and the Australian Code for the Responsible Conduct of Research requires researchers to make data available to others where possible and appropriate (R22). By making data available for reuse in future research, researchers increase the potential benefits of collecting data, and reduce the future burden on participant groups. Ethics applications generally involve some level of information about a researcher’s plans for managing their research data, and both the ethics process and the conditions of the approval can impact a researcher’s thinking about and approach to research data management.
At the outset, it is important that researchers understand that their responsibilities around research data management go beyond what is required for approval by a research ethics committee. A research ethics committee can assess the overall ethical appropriateness of research plans, but this assessment takes place within the context of the broader requirements of the university, with the assumption that the researcher understands and will abide by those requirements. There is a risk that researchers may view ethics approval as the primary test of the appropriateness of all aspects of their research plans, including data management. If researchers view ethics as the step at which the university checks that they are “doing things properly”, then they may have the attitude that once they have received ethics approval, they are “done.”
Ethics review is typically concerned with data management at the higher level – for instance, ensuring that risks associated with the data have been identified, and that the data will be stored securely and only accessed by the appropriate people and in line with participant consent. The level of detail considered within the ethics application varies, and is unlikely to cover every aspect of research data management. It is important, however, that researchers carefully consider all aspects of their research data management to ensure that data is handled safely and appropriately. The National Statement on Ethical Conduct in Human Research requires researchers to create a data management plan, and these plans are also strongly encouraged in the guidance supporting the Australian Code for the Responsible Conduct of Research. To assist researchers with this, most institutions provide guidance, templates or tools for research data management planning. These take researchers through a more in-depth planning process, often with links to institutional policy and procedure.
However, for those researchers who are not required to submit a research data management plan as part of university processes, ethics approval may be the only formal process through which they consider the management of their research data. Researchers are conscientious and want to conduct their research in accordance with their obligations , but given the very high demands on their time they can rely on external prompts to do so. Ethics approval acts as a major external prompt, requiring researchers to think about the questions covered in the application. Once this is done, other priorities can emerge. It may be helpful to explicitly prompt researchers to complete a separate research data management plan when asking questions about data management in the ethics application.
Some institutions do require more detailed information about data management in the ethics application (or even the submission of a full data management plan alongside the application), but this comes with its own difficulties. Research ethics committee members are selected for their knowledge and expertise, but they are not necessarily experts in the details of research data management, a complex and fast-moving field that involves an understanding of cybersecurity, national and international regulatory requirements, institutional policy, and community-agreed best practice. This lack of expertise can make it difficult for ethics committees to be confident in their judgements. For instance, a committee might seek to ensure that identifiable data about participants is kept securely, and that only data that cannot identify participants be shared with external project collaborators. However, assessing whether participants can be re-identified, now or in the future, from a given dataset can be complex, even when obvious identifying variables have been removed or transformed. Without a strong understanding of current best practice, it can be hard to determine whether a researcher’s plans for their data present an unacceptable risk.
Committee members may also vary in their knowledge of local policy. Each university will have their own requirements for research data management, including guidance on which university platforms are appropriate for securely holding sensitive data. External committee members may have difficulty even accessing these policies, and internal committee members may not be fully familiar with them. Policy and procedure will change with the emergence of new requirements, systems and technologies, and as best practice shifts. For instance, where in the past it was more common to dispose of data after the minimum retention period, university policy is now more likely to encourage researchers to archive data that could be reused for future benefit. Some universities reported that their ethics processes were lagging behind in supporting this aspect of best practice, with application forms and committee decisions defaulting to the disposal of data.
Researcher expertise in research data management also varies widely. This can create a disconnect, where both researchers and committee members expect the other to have the expertise to make decisions about the appropriateness of research data management. A research ethics committee may assess plans based on the assumption that the researcher, the expert in the field of study, has correctly classified the sensitivity of their data, while the researcher assumes that if they have got the classification wrong then the committee will alert them. To close this gap, both researchers and research ethics committee members require training and guidance on research data management. Research ethics committees could benefit from self-paced training materials that fit with demands on members’ schedules and the continuous cycle of induction of new members. It could also be useful to provide guides which summarise the institution’s policies and requirements around research data management, particularly for external members who may have difficulty locating these materials.
One of the most valuable resources for research ethics committees in making decisions that relate to research data management is the existing expertise of data management professionals at the institution. Cybersecurity, IT, eResearch and library staff, among others, work directly with data management daily and are well placed to advise on best practice and assess researcher plans. Depending on structure and resourcing, universities have found different ways to inject this expertise into the ethics process. One approach is to institute formal processes which allow committees to refer a research ethics application to cybersecurity for review when it is unclear if planned data management solutions are sufficiently secure. Another is to build a relationship between research data experts, ethics administrators and committee members through a period of face-to-face consultation. Where resources allow, placing research data experts on the research ethics committee itself allows excellent review of the appropriateness of data management planning. Whichever of these approaches is taken, working directly with those who have a deep understanding of research data management helps research ethics committees to be confident that they are assisting researchers to manage their data ethically and appropriately.
With thanks to the universities that participated in the Institutional Underpinnings program, and the experts who took the time to discuss these issues further.
FONG, C., HAYTHORNTHWAITE, A., RIDGEWAY, K. (2022) Are HRECs our sensitive data advocates? How working with HRECs can improve research data management outcomes. Presented at eResearch Australasia 2022 (https://conference.eresearch.edu.au/2022/08/are-hrecs-our-sensitive-data-advocates-how-working-with-hrecs-can-improve-research-data-management-outcomes/)
NEISH, P.; DAVIS, N. (2022): Institutional Underpinnings – Ethics and Data Discussion Paper. University of Melbourne. Report. https://doi.org/10.26188/20500101.v2
This post may be cited as:
Burton, N. (14 March 2023) Ethics Committees and Research Data Management Research Ethics Monthly: https://ahrecs.com/ethics-committees-and-research-data-management/