A A A
OPB logo

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

6. Dedicated Consultants to Offer One-to-One Support with Data

What if you have slightly more resources available and would like to effectively engage with researchers across all disciplines? We recommend you consider hiring dedicated domain experts who could offer in-depth, one-to-one support to your researchers.

We collected three slightly different case studies illustrating this approach:

  • Data Stewards at TU Delft;
  • Informatics Lab at Virginia Tech;
  • Data Managers at Utrecht University.

Subject-Specific Consultants Are an Add-On to ‘Traditional’ RDM Support at Large Institutions

In all of these approaches, people who provide the consultancy support are hired full-time to do this job. In all cases, they also have a research background to facilitate their interaction with the research community. All of our examples come from large, research-intensive universities, which have more than 10 people providing centralised Research Data Management (RDM) support in addition to these domain-specific consultants.

‘Show Me the Money’

One differentiating factor among the three cases is the business model. While both TU Delft and Virginia Tech pay for the salaries of their domain consultants from faculty budgets, at Utrecht University the salary of data managers is covered by research grants. This has implications for the feasibility of implementation elsewhere. The TU Delft and Virginia Tech models are more challenging to implement, because of the need to allocate central budget for the positions. The Utrecht University model offers more flexibility and scalability: the greater the demand, the more projects ask for data management support, and the more money becomes available to hire data managers.

All three cases have slightly different flavours and twists to them. So without any further spoilers, enjoy our three picks!

6.1. Data Stewards at TU Delft: A Reality Check for Disciplinary RDM

Author: Yan Wang

Contributor: Alastair Dunning

© Yan Wang and Alastair Dunning, CC BY 4.0 https://doi.org/10.11647/OBP.0185.18

TU Delft employs former researchers as Data Stewards within each faculty, creating a full-time local contact for RDM advice who engage with researchers and inform them about other institutional support services and resources.

Table 6.1, CC BY 4.0.

Fostering cultural change is the objective of the Data Stewardship program. Delft University of Technology (TU Delft) in the Netherlands were bold: they hired 8 researchers to act as Data Stewards at each faculty. These are professional Research Data Management (RDM) specialists tasked with improving daily data management practices within their research communities. ‘Our message is simple. If researchers have any questions about their research data, the Data Steward is their go-to person. Researchers are there to do research; they can’t be expected to know everything about the latest tools available, or about all the nitty gritty details of policies and regulations. Data Stewards serve as bridges between the researchers and all other research support services, such as the library, ethics committee, ICT (Information, Communication and Technology), privacy and legal teams,’ says Yan Wang, the Data Steward at the Faculty of Architecture and Built Environment. The daily job of the Data Steward is therefore to respond to researchers’ requests, advise them, and promote good RDM practices.

Job Definition: Have Disciplinary Expertise in Data Management, Take Initiative and Be a People Person

All Data Stewards are former researchers. They have a PhD degree, or equivalent experience to match the research background of the faculties they work for. This allows them to fit in with the faculty culture and develop discipline-specific data policies.

Being the only RDM person serving the entire faculty presents challenges. ‘You need to take a lot of initiative to define the job, to reach out to people and make yourself visible. RDM is still a new subject, and it is challenging to help researchers, especially those who are not aware of the benefits of good RDM practices. The only way to raise awareness is to meet them, talk to them, get to know their work and make them understand what the benefits of good data management practices are,’ explains Yan.

This is why Data Stewards need to have exceptional communication skills and enjoy working with people. ‘Interpersonal skills are key to the success of the Data Steward,’ reflects Alastair Dunning, the Head of Research Data Services at TU Delft. Alastair was one of the founders of the Data Stewardship programme. ‘Once we connect with researchers, the interactions are typically very positive and allow development of strong relationships and incremental improvement of data management practices,’ adds Yan.

Fig. 6.1 Software carpentry workshop organized by Data Stewards. © Yan Wang / TU Delft, CC BY 4.0.

Coordination Is Crucial to Create Operational Synergy

Data Stewards are embedded in faculties and centrally coordinated by the library-based Data Stewardship Coordinator. In order to provide comprehensive research support, operational synergy among all support teams at the university is paramount. Good RDM practices put new requirements on the workflows where different teams are involved. The Data Stewardship Coordinator plays an important role in bringing the different teams together. The Coordinator constantly steers the communication and facilitates joint efforts.

Institutional Support Is Needed for Implementation

In response to emerging trends, new positions are needed at research institutions. This is not a common practice, and needs support from senior management willing to take risks and provide adequate investment to allow innovations and developments. ‘Perhaps we should focus not so much on having the perfect technical infrastructure, but on whether we have the right people. People are key drivers of cultural change and that’s the essence of our data stewardship initiative,’ concludes Alastair.

6.2. Cultural Change Happens One Person at a Time: Informatics Lab at Virginia Tech

Author: Marta Teperek

Contributor: Jonathan Petters

© Marta Teperek and Jonathan Petters, CC BY 4.0 https://doi.org/10.11647/OBP.0185.19

Virginia Tech employs researcher-consultants to provide expert support within specific domains and build good practice through ongoing partnerships with researchers.

Table 6.2, CC BY 4.0.

‘To a large extent, researchers are not interested in data management planning. Apart from the fact that they have to do it for grant proposals, they’re generally not interested in sharing data. They want their problems to be solved: “my workflow is really inefficient and I wish it were better”, or, “I would really love to use this new software, but I can’t figure out how to get it to work”. So our aim with the Informatics Lab is to help researchers where they want help. While helping them we have the opportunity to talk to them about data management planning and data sharing,’ explains Jonathan Petters, Data Management Consultant and Curation Services Coordinator at Virginia Tech.

Fig. 6.2.1 Jonathan Petters, Data Management Consultant and Curation Services Coordinator at Virginia Tech © Jonathan Petters, CC BY 4.0.

Domain-Specific Consultants at the Informatics Lab

The Informatics Lab was established as part of the Data Services unit with the goal of helping researchers across all disciplines to deal efficiently with their data. The team consists of five people: four consultants and one coordinator.1 All consultants have domain knowledge to help them develop deeper interactions and relationships with researchers. ‘Finding the right person for the position is a challenge,’ says Jonathan. ‘You need people who have a good understanding of the research lifecycle and of data management, but also the ability to think at a higher level than just one research project. You want a combination of both depth and a higher, broader view.’

Jonathan also explains that in addition to all the research requirements, it is crucial for the consultants to have a genuine interest in offering support: ‘If you’re not interested in helping out other people, you’re not the right person. A good way to test this is to make sure that people to which the services will be provided are involved in the hiring committee.’

Research Background — A Double-Edged Sword?

The informatics consultants are employed as permanent faculty members, meaning that they are researchers in their own right. They are expected to maintain a research portfolio and to publish. ‘Actively doing research is not in direct opposition to being a good service provider, but I think that there’s a bit of tension. Some of our consultants feel they are here to help. Others focus more on their research and do some consultancy on the side,’ reflects Jonathan. ‘But the benefit is that by doing their own research, it’s even easier for them to get out and talk to other researchers. They form organic networks through shared interests and connections, and that’s how mutual awareness and trust is created. I could talk to a biomedical researcher about human subjects’ data because I know these issues pretty well. But if I say that my background is meteorology, they may say, okay, so you know something about research, but that’s from a different domain. Having somebody who’s got that background brings credence.’

Fig. 6.2.2 Informatics Lab at work. © Ann Brown/Virginia Tech, CC BY 4.0.2

Five Full-Time Employees Are Expensive — Are They Worth the Investment?

The group has had over 200 consultations so far. They have a big body of evidence that they will eventually look into, but Jonathan prefers not to rely solely on quantitative feedback: ‘Everybody was saying they were really happy. So unless people are willing to be upfront and critical with you, you’re not likely to receive helpful feedback. Nobody is going to say: “I know you tried to help me for 6 hours, but in fact you made it worse”.’ Jonathan prefers to focus on long-term effects. ‘When our consultants talk to researchers about the kind of data they produce, about all those different proteins and complex molecules, for example, they help researchers realise that it’s actually all data. They get them to think about their research in a different way. This opens up an opportunity to have a conversation about changing their practice.’ Jonathan reflects that when they help a researcher, they get a recurring client who views the library as a partner. ‘It’s a one person to one person thing and it leads to slow cultural change. We also help research students and that’s going to have an impact on them moving forward. Whether they are all going to become tenure-track professors or not, it does contribute to cultural change.’

6.3. Ever Heard of Five-Legged Sheep? Data Managers at Utrecht University Give Researchers a Leg-Up!

Author: Iza Witkowska

Contributor: Martine Pronk

© Iza Witkowska and Martine Pronk, CC BY 4.0 https://doi.org/10.11647/OBP.0185.20

Utrecht University employs a pool of data managers who can be flexibly hired and embedded within research teams for short or long periods, providing highly targeted RDM support and allowing researchers to concentrate on research.

Table 6.3, CC BY 4.0.

‘Demands in the various phases of the data cycle are so diverse that finding all expertise in one person is comparable to searching for the proverbial “five-legged sheep”,’ says Martine Pronk, the head of Academic Services of the Utrecht University (UU) Library.

The ‘five-legged sheep’ is an old Dutch expression, referring to somebody who needs to be unreasonably versatile. And sadly, this is what is expected from the twenty-first-century researcher.3 Transparent and well-documented data management is one of many tasks that researchers now need to add to their already heavy workload.

The UU flexible data managers’ pool takes the technical aspects of managing data away from overwhelmed researchers. A data manager can be hired from the library and embedded in a research project part-time or full-time, either short-term for small, specific tasks, such as support in writing a data management plan, or for longer periods, for example to set up a data flow and help to manage the data collection over an entire project. In this way, expertise that is developed stays within the university, in contrast to the scenario of hiring an external (and most likely expensive) expert for the same job.

At UU the data manager service was set up in response to a request for expertise. In 2017 the program director of an ongoing large-scale study with multiple partners, the YOUth cohort study,4 requested a skilled professional to help with managing data. The library responded enthusiastically to this request and the service of the embedded data manager was born. Reflecting on the success of this project, it was expected that more of these projects would be initiated at the UU (and elsewhere). The popularity of this service has since increased and it has been promoted as vital by satisfied customers5 to their peers.

‘The data managers’ pool fits well in the traditional values of the library to make research output findable, available, accessible and re-usable. What’s new is that in response to the changing needs of the scientific community and digital innovations, the library and its employees move closer to the researchers,’ says Martine.

The best way to start this service is by having large, data-driven, complex projects or programmes with multiple (inter)national stakeholders. These projects are more likely to have enough funding and to require high quality data management support to manage all the data flows. It also helps to have a well-developed Research Data Management (RDM) network within the university to promote the service among the researchers and help identify researchers and projects that could benefit from it.

The Secret Ingredients Are People

The perfect candidate for the role of data manager should be someone who is able to act as an advisor, project manager and leader, as well as contributing technical skills in RDM. He or she needs to be proactive, flexible, show initiative and be able to work effectively within, and contribute to, a positive team environment. They also need to be capable of building productive networks both internally and externally. And of course, they need to have both passion and experience with managing data and the ability to communicate their expertise to others.

For data managers, having a PhD might be helpful to understand the research environment, but is not required. What is necessary is the ability to communicate effectively with researchers. Because this function is new and the field of data management is dynamic, data managers have a unique opportunity to develop and customize their role according to the needs of the scientific community and their own professional interests.

Fig. 6.3 Data managers Ron Scholten and Danny de Koning at the KinderKennisCentrum of Utrecht University. © Annemiek van der Kuil | PhotoA.nl, all rights reserved.

The main challenge for the service is acquiring new projects and doing so in a timely manner, in order to have enough projects on the go at any given time. The pool and individual data managers have targets and they need to meet them: currently 65% of the collective FTEs (Full-Time Equivalents) need to be outsourced to projects. To achieve this goal, data managers need to look continuously for projects and use their networks to increase the visibility of the service.

Although it’s a relatively new service, there are two clear indicators showing that this service is well received by the scientific community and promotes good data management practices. ‘Firstly, interest, demand and appreciation for this service is growing, and researchers themselves promote this service to their colleagues. Secondly, data managers act as ambassadors of FAIR (Findable, Accessible, Interoperable, Re-usable) and open data. They educate and advise researchers on good data management practices, including how to archive or publish data and metadata standards. More and more datasets get published at the UU,’ says Martine.


1 Informatics Lab at Virginia Tech, https://informaticslab.lib.vt.edu/

2 Virginia Tech News, ‘University Libraries Has Expertise, Resources to Help Faculty Overcome Data Challenges’, where the photo was originally published, https://vtnews.vt.edu/articles/2019/05/univlib-datasalvage-miller.html

3 ‘Reflections on Research Assessment for Researcher Recruitment and Career Progression — talking while acting? ‘ (blog post), https://openworking.wordpress.com/2019/05/20/reflections-on-research-assessment-for-researcher-recruitment-and-career-progression-talking-while-acting/