OPB logo

Engaging Researchers with Data Management: The Cookbook
More info and resources at: https://www.openbookpublishers.com/product/1080

2. Finding Triggers for Engagement

Despite the benefits of data management planning for the researcher, many still regard it as an administrative burden. Interestingly, some institutions were able to turn data management planning into an opportunity to engage researchers in discussions about data.

This chapter looks at several case studies where workflows have been designed to bring about interaction and engagement at key moments in the research process.

Being able to work in collaboration with other support or management units within the university was a key factor in each of these.

2.1. Taking Advantage of Existing Administrative Systems: MRC/CSO Social and Public Health Sciences Unit, University of Glasgow

Authors: Joanne Yeomans, Iza Witkowska

Contributor: Mary-Kate Hannah

© Yeomans, Witkowsk and Hannah, CC BY 4.0 https://doi.org/10.11647/OBP.0185.03

Piggybacking onto an existing system for approving research proposals, the Health Sciences Unit at the University of Glasgow automatically contacts researchers that might need RDM support at the beginning of their projects, and then follows up throughout the project’s lifespan.

Table 2.1, CC BY 4.0.

Engagement as Early as Possible — The WHY

In the MRC/CSO Social and Public Health Sciences Unit at the University of Glasgow,1 the department representatives decided to use an existing administrative reporting system to help engage with researchers about data management issues and to manage requests coming to the IT (Information Technology) and other support staff offices.

Part of the reason we set up this system is that people would apply for grants, and then when the research started they’d go to the support staff and ask: can you help me with transcribing, can you help me with fieldwork or whatever else was needed, and the support staff representative would say: we don’t have this in our diary, we have two other surveys happening at the moment, so we can’t do this, we need warning that these things are going to happen. So now, because it’s reported in advance in the system, they can plan, they can take on new staff, anything that is needed. — Mary-Kate Hannah, Data Scientist in the Unit.

How Early Is Early? The HOW

Whenever a project is initiated, a researcher has to fill in and submit an online form. The research proposal is considered by the ‘Portfolio Group’ which checks that the topic is in line with the unit’s focus and identifies what resources might be needed within the department, whether space for staff members, IT facilities, and so on.

The Portfolio Group consists of senior and experienced research staff, and senior representatives of all the different research programmes. Representatives from various support offices also sit in on the meetings. The group considers the proposal and, among other things, identifies whether there is data collection or data creation planned and whether data will be stored at Glasgow University. If so, they tick a box in the form that indicates a data management plan (DMP) is required and this triggers an automatic email to Mary-Kate and to the submitter so that they can follow this up. Without Mary-Kate signing off on the completion of a satisfactory DMP, the researcher cannot move forward with their grant application or their research.

The automated email starts the process of putting Mary-Kate in touch with the researcher to assist with the writing of the plan. This happens regardless of whether a funder requires a DMP or not: if the research is going to generate data that will be stored at Glasgow University, then the department itself still requires a DMP to get things right from the start.

Fig. 2.1.1 Mary-Kate Hannah helping a researcher to complete a DMP at the MRC/CSO Social and Public Health Sciences Unit, University of Glasgow. Photograph by Enni Pulkkinen, 2019, CC BY 4.0.

Benefits for the Researcher — The WHAT

So, what happens after the sending of the automated email saying that a DMP is needed? The next step involves Mary-Kate sending a customized email and flagging up resources such as the pre-filled DMP template and the DMPonline tool, as well as offering support on any other Research Data Management (RDM) relevant aspect (for example, handling personal data). She also refers researchers to a recorded presentation about RDM consisting of PowerPoint slides with recorded voiceover. This personal touch is definitely the key factor in making the implementation of this process successful.

The RDM support doesn’t end there. Once a project begins, a study or trial master file2 is set up on the network drive and customized for the project team. It has standard folders for storing common administrative documentation, such as grant application and legal documents, and includes a folder for data management. This generic folder structure was developed after looking at and reviewing many studies at the department. ‘Researchers and support staff are very happy with this; it saves time as they don’t have to think about this themselves,’ says Mary-Kate.

Fig. 2.1.2 Mary-Kate Hannah delivering a training session on data management planning at the MRC/CSO Social and Public Health Sciences Unit, University of Glasgow. Photograph by Enni Pulkkinen, 2019, CC BY 4.0.

Looking Back, Does It Work?

Yes, it does! People are getting used to the idea of planning their data management: because they did it for their last project, they are expecting it for their next project. They have often reported that the process of writing the DMP has been helpful and that they have found conversations with the research data management advisor to be useful.

Having said that, there is still room for improvement in the system to save researchers’ time. For example, researchers need to write similar details in different forms for their grant application, their ethical review request, and their DMP. These forms are delivered at different times and the procedural timing could be better optimized.

Sometimes the detailed information requested in the DMP comes too late, for example, ‘the researcher might have had ethical approval for their data collection consent form which did not contain data-sharing information, then they start to write their data management plan and realize that they will have to make an amendment to their consent form and ask for an amendment to their ethics application,’ says Mary-Kate.

What is clear though, is that embedding a requirement for data management planning directly in the unit’s authorization process is crucial in getting researchers to think about their data management at an early stage, and in putting RDM support staff in contact with researchers right from the start.

2.2. Engaging with Researchers through Data Management Planning at the University of Manchester

Author: Joanne Yeomans

Contributors: Rosie Higman, Christopher Gibson

© Yeomans, Higman and Gibson, CC BY 4.0 https://doi.org/10.11647/OBP.0185.04

The University of Manchester illustrates how careful design of DMP templates and DMP policies allows staff to effectively engage with researchers through DMP review.

Table 2.2, CC BY 4.0.

Requiring an approved Data Management Plan (DMP) before allowing a research project to begin might work for smaller units within a university, but it might not scale up across a large research-intensive university. The University of Manchester has found a way to design their DMP template and DMP requirement process so that the library support team can potentially engage with all researchers at an early stage of their research. Focusing the engagement around the writing of the DMP means that they can offer advice when a researcher is beginning their project, but they can also learn first-hand whether the university’s data management policies are practical.

At the University of Manchester, a DMP is required when applying for funding, ethics approval and/or IT storage. ‘Between these you encompass most research at the university,’ thinks Rosie Higman, Research Services Librarian. ‘I’m not saying we’ve got it perfect but at least in theory, we’re going to cover most projects through these routes.’

The University has DMP templates in the DMPonline3 system that ask university-specific questions in the first section (see Fig. 2.2.1). These include questions about storage needs and whether the research is handling personal data using a tick-box format so that the questions are easier to answer. The University’s Information Governance Office uses this part of the form as the asset register in accordance with the European General Data Protection Regulation (GDPR).

Fig. 2.2.1 Overview of the University of Manchester DMPonline template showing the Manchester-specific questions. University of Manchester Library, CC BY 4.0.

The Research Data Management team (Fig. 2.2.2) in Research Services at the university library check this first section of every DMP and give feedback to the submitter.

Fig 2.2.2 University of Manchester Research Data Management team. From left to right: Jess Napthine-Hodgkinson (Research Services Officer); Clare Liggins (Research Services Librarian); Chris Gibson (Research Services Librarian). Rosie Higman has since started a new position at the University of Sheffield. University of Manchester Library, CC BY 4.0.

What Do You Learn from Checking So Many DMPs?

‘You can certainly tell which DMPs are from people who have been to our training,’ Rosie is pleased to point out. ‘It’s helped us work out where we have lots of gaps, where the policy is unrealistic or the procedures are unsupportable.’ She gives an example: ‘There’s a university procedure for when a researcher makes a recording of a participant; it’s clear and well written and has been around for some time, but it suggests that every researcher should have access to an encrypted recording device and the university is only just working out what the cost of that would be.’

When talking to researchers about a DMP, you are therefore sometimes challenged, ‘how do you do that in practice?’ In the case of encrypted recordings, the Research Data Management (RDM) team, with the help of information governance and IT (Information Technology), has been able to draw up a list of practical steps a researcher can take, but this has raised difficult questions about whether the policy should stand. ‘It’s making our services more responsive to what researchers want,’ concludes Rosie.

Avoiding Being a Victim of Your Own Success

Almost 20% of the submitters request a more detailed review of their DMP and most of these requests are from researchers dealing with personal data or ethical permissions.

Rosie and her team have some standardised answers, which an officer tailors to each case before drafting a response. One of three librarians will then review the officer’s comments and enhance them with more discipline-specific suggestions. They aim to treat each plan in under an hour. ‘We already spend significant time on this and every week meet to discuss the DMP review requests that have come in and how we can balance them with our other work priorities. If the demand increases, we’re not yet sure how to address this,’ admits Rosie.

One technique that they believe will save review time is to remove the option to allow free text and instead offer tick-boxes in answer to fixed questions. The use of pre-drafted comments for responses has also helped, but the time it takes to review a single DMP is still a challenge.

The use of DMPonline started in Manchester in 2018, when the European GDPR also came into effect. Although all researchers are required by university policy4 to write a DMP, it’s not clear what proportion of researchers are doing so. The university also has a system for tracking student progress in general, which requires students to have a conversation with their supervisor, during which there is a prompt to check that the student has a DMP. ‘This is a good start, but obviously carries the risk that if a supervisor does not care about data management then students will not create a DMP,’ says Rosie.

With just over a year of experience in using DMPonline in this way, the library team thinks it is a good time to review the level of compliance. They will do this by checking the institutional records to identify the proportion of research projects that have a DMP and expressing this as a percentage for the university, faculties and schools. The results, expected in late 2019, will be interesting to compare differences in behaviour and should give the team some idea of how demand for reviews might increase.

2.3. Timing Is Everything When It Comes to Engaging with Researchers at the University of Technology Sydney

Author: Iza Witkowska

Contributors: Wendy Liu, Duncan Loxton, Elizabeth Stokes, Sharyn Wise

© Witkowska, Liu, Loxton, Stokes and Wise, CC BY 4.0 https://doi.org/10.11647/OBP.0185.05

At the University of Sydney, support staff provide grant recipients with ‘stub’ DMPs and interviews at the right time to maximize researcher engagement.

Table 2.3, CC BY 4.0.

At the University of Technology Sydney, the eResearch Unit5 in the Central IT (Information Technology) Division and the Library’s Research Data Team6 collaboratively approach recipients of major research grants and offer them a 45-minute interview to provide data management support and create a data management plan (DMP). The aim of this activity is to help grant recipients to comply with the data management plan policy from Australia’s major research funders and to simultaneously engage them with discussions about research data.

Many Research Data Management (RDM) support services within universities and research institutions do this, so what makes the work by the team from the University of Technology Sydney so successful and noteworthy? Well, it’s all in the details.

First, they come to researchers with a ‘stub’ DMP: a pre-filled DMP based on the abstract of the funded grant application. This advance work helps to make things run more smoothly and means they can structure the interview around whether the draft plan accurately characterises the researchers’ data management activities and requirements. And why does this approach work? Well, as the saying goes ‘you catch more flies with honey than with vinegar’. The stub DMP is the ‘honey’ because it takes researchers one step closer to meeting funders’ requirements. The expected outcome of the interview is not to have a completed DMP, but to have started a conversation about data management.

Second, persistence. It’s not always easy to get the lead investigator to respond to the first contact, but our colleagues from Sydney don’t give up. They repeatedly attempt to schedule an interview, and will approach more junior researchers on the project, especially those responsible for data curation/custodianship, if the lead investigator remains unavailable.

Third, their timing is right. These interviews target research teams at the right point in the project cycle to make data management decisions. They also provide an immediate connection to eResearch support if complex software or computational infrastructure is required.

There are benefits on both sides. Researchers engage with research data, and gain awareness of RDM infrastructure and the support available at the university before they need it. The provision of appropriate data storage solutions, software or other infrastructure for their research projects is guaranteed. Policy compliance becomes less of a hurdle, and the increase in collaboration between service units within the university helps to break down institutional silos. Librarians can demystify research data management practices for researchers in a friendly way, while gaining a deeper understanding of specific data management requirements.

The good news is that any organisation able to provide a DMP tool (in this case, Stash,7 a home-grown service integrated into the research management system) and build communication between IT infrastructure and Library/RDM services, can implement a similar initiative. Good communication channels and the ability to provide a swift follow-up are also essential. In order to achieve this, it’s helpful to have a coordinator in place, especially someone familiar with the available IT infrastructure.

To take this service to the next level, this activity can be linked to broader institutional campaigns surrounding academic integrity, raising its profile within the university. Other options are to strengthen collaboration with other research support offices and collect evidence that DMPs improve data management practices, for example, by conducting user satisfaction surveys. To secure the project in the longer term, it is also important to document and communicate its success to senior administration.

1 MRC/CSO Social and Public Health Sciences Unit at the University of Glasgow receives joint core funding from the Medical Research Council (MRC) and the Scottish Government Chief Scientist Office (CSO).

3 DMPonline tool, https://dmponline.dcc.ac.uk

4 University of Manchester Research Data Management Policy, http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=33802