Dealing with Data 2019 (January 2020): Collaboration Across the Nations

Picture the scene: A cold January day, the wind blowing the scarves of the passers-by through the large windows of the Informatics Forum meeting room. The group inside listens, takes notes, tweets, and asks questions of the speakers, representing a range of disciplines across the University…

Dealing with Data is an annual event hosted by the Research Data Service. Its aim is to engage the University community of researchers and support professionals around a theme, to share success stories and challenges in the myriad, everyday issues involved with data-driven research. The theme this year reflected the difficulty of managing research data in large, collaborative projects. Due to industrial action, the original November event was postponed to January. Around a hundred researchers – staff and students – participated, along with support staff who gave lightning talks about research-focused services. Full presentations and videos are now available.

So Benjamin Bach, our keynote speaker, inspired us with state of the art data visualisation software and techniques for both exploration and presentation. But he also illustrated the difficulties of portraying all of the data in all of its facets of a rich dataset, and the consequences of making necessary choices for its interpretation.
The first session began with Tamar Israeli’s study of researchers’ use of collaborative and institutional tools showed the challenges of making local infrastructure user friendly enough to attract new users familiar with slick cloud-based services. Then Mark Lawson demonstrated his ingenuous ‘ethical hacking’ to piece together a set of APIs to create a research workflow for samples and images for histology research. Minhong Wang conveyed a higher level view of data management focused not just on data-driven, but knowledge-driven phenotyping.

Next were the lively lightning talks, in which Mike Wallis of Research Services warned of a new Digital Dark Age, and David Creighton-Offord spoke of the dillemmas in Information Security user support where shiny doesn’t always equal safe. Lisa Otty spoke of innovative training and text mining projects bringing data science to the Humanities, and Rory MacNeil demonstrated how the RSpace electronic lab notebook can connect to a host of popular open science tools.

Following a lively lunch with chat between delegates and with hosts of the service exhibitions, Alex Hutchison showed a highly programmatic view of data management and ethics control from the UNICEF collaboration, in collecting and analysing real world data about children in need. Caileen Gallagher offered a case study of how food courier data could be used to empower workers. Sanja Badanjak shared her data integration problems of peace agreements around the world, conveying both innovative solutions and time-consuming workarounds.

In the final session Edward Wallace brought in the Edinburgh Carpentries to the rescue of poor data skills within Biological Sciences and the wider University – itself a great example of cross-community collaboration building a community of trainers. Gillian Raab showed us how any data problem however intractable can be solved by resourcefulness and determination, making use of DataShield for multi-party computation when datasets are too sensitive to be shared. Johnny Hay and Tomasz Zielinski demo’d their Plasmo ‘boutique repository’ for plant-systems biology modelling and Holly Tibble described tackling an international collaboration in linking administrative datasets via ‘ridiculously detailed’ statistical analysis plans. Representing the Research Data Service, I wrapped up proceedings with some of these very observations.
Both presentations and videos are available.

Welcome

  • Jeremy Upton, Director of Library and University Collections. [Presentation]

Keynote

  • Data Visualization for Exploration and Presentation, Prof. Benjamin Bach. Lecturer in Design Informatics and Visualization. [Presentation] [Slides]

Session 1 – Chair: Theo Andrew

  • “Data Something”: Assessing Tools, Services and Barriers for Research Data Collaboration at the University of Edinburgh – a small-scale study carried out by Dr Tamar Israeli with support from the Research Data Support team. Robin Rice – Data Librarian & Head of Research Data Support Services. [Presentation] [Slides]
  • Integrated secure web application to deliver centralised management of research samples, histology services and imaging data. Mark Lawson, Data & Project Manager, MRC Centre for Reproductive Health, QMRI. [Presentation] [Slides]
  • Building the Knowledge Graph for UK Health Data Science Minhong Wang et. al, Deanery of Molecular, Genetic and Population Health Sciences. [Presentation] [Slides]

Session 2 – Chair: Kerry Miller

  • The Data Opportunities & Challenges when Collaborating across Organisations
    Alex Hutchison, Delivery Director – Data for Children Collaborative with UNICEF. [Presentation] [Slides]
  • Restoring Gig Workers to Power: Personal Data Portability, Supply of Digital Content and Free Flow of Data in the European Data Economy. Cailean Gallagher, Scottish Trades Union Congress, & St Andrews University Institute of Intellectual History. [Presentation] [Slides]
  • Dealing with data in peace and conflict research. Sanja Badanjak, Postdoctoral Research Fellow, School of Law. [Presentation] [Slides]

Session 3 – Chair: Robin Rice

  • Bringing researchers to data: computing skills training with Edinburgh Carpentries.
    Edward Wallace, Sir Henry Dale Fellow, Institute of Cell Biology. [Presentation] [Slides]
  • Running an analysis of combined data when the individual records cannot be combined. Gillian M Raab and Chris Dibben, Scottish centre for Administrative Data Research. [Presentation] [Slides]
  • The grant is dead, long live the data. Johnny Hay and Tomasz Zieliński, School of Biology, University of Edinburgh. [Presentation] [Slides]
  • International collaborations using linked administrative data: Lessons from the MARIC study. Holly Tibble, Usher Institute, University of Edinburgh. [Presentation] [Slides]

Robin Rice
Data Librarian and Head, Research Data Support
Library & University Collections

Share

A visit from the data jungle: My internship in research data management

This is a guest post from Dr. Tamar Israeli, who completed a work/study internship with the Research Data Support team last Autumn. A link to her report is available below.

Recently, there has been a rumor in Israel that research data should be managed. As a librarian and information specialist working in an academic institution, I decided to check if this was true.

When looking for a place for an internship on the role of the library in research data management (RDM), I was happy to find out that the University of Edinburgh RDM support team has a good reputation. I remember enjoying very much my visit to Edinburgh 30 years ago so I was very happy to get Robin Rice & Martin Donnelly’s kind invitation so I could boldly go where… I had already been before.

During September 2019, I worked with the RDM support team, attended some of the staff meetings and participated in one of the RDM trainings.  As part of my internship we carried out a small scale study. The purpose of the study was mainly to understand what are the barriers that prevent researchers from using tools and services provided to them by the university when collaborating with data.

For that purpose, I interviewed six researchers from different schools and disciplines. The researchers were open and cooperative and the interviews were very interesting and insightful. If you’d like to learn about the way researchers collaborate and what influences their decision to use a particular tool or service, here is a link to our report: http://dx.doi.org/10.7488/era/2

Many thanks to the support team for their invitation and warm hospitality. It was one of the most pleasant months of my life.

Tamar Israeli
Librarian and information specialist
Western Galilee College

Share

New research data management tool on one-year trial: protocols.io

Information Services aims to offer a research data service that meets most of the data lifecycle needs of the majority of UoE researchers without interfering with their freedom to choose tools and technologies which suit their work. In some cases cloud tools that are free to individual users are offered commercially as enterprise versions, allowing groups of researchers (such as lab groups) to work together efficiently.

The service’s steering group has agreed a set of criteria to apply when a tool is put forward by a research group for adoption. The criteria were developed after our two-year trial of the electronic lab notebook software, RSpace, and have been most recently applied to protocols.io. The protocols.io trial begins this month and will run for one year. An evaluation will determine whether to continue the enterprise subscription and how to fund it.

protocols.io is an online platform for the creation, management, and sharing of research protocols or methods. Users can create new protocols within the system, or upload existing methods and digitise them. Those with access to a protocol can then update, annotate, or fork it so that it can be continually improved and developed. There is interoperability with Github and RSpace, and long-term preservation of protocols through CLOCKSS.

Users can publish their protocol(s) making them freely available for others to use and cite or, with the enterprise version, keep them private. The tool supports the Open Science / Open Research agenda by helping to ensure that methods used to produce data and publications are made available, assisting with reproducibility.

Subscribing to the University plan will allow research groups to organize their methods and ensures that knowledge is not lost as trainees graduate and postdoctoral students move on. There are currently over 70 University of Edinburgh researchers registered to use protocols.io. You may follow these instructions to move your current protocols.io account to the premium university version. For more information contact data-support@ed.ac.uk.

Kerry Miller and Robin Rice
Research Data Support team

Share

Collaborating on data in a modern way

Between mid-September and mid-October, the Research Data Support team hosted an international visitor. Dr Tamar Israeli, a librarian from Western Galilee College in Israel, spent four weeks in Edinburgh to increase her experience and understanding around research data management. As part of this visit, Tamar conducted a study into our researchers’ collaborative requirements, and how well our existing tools and services meet their needs. Tamar’s PhD thesis was on the topic of file sharing, and she has recently published another study on information loss in Behaviour & Information Technology: “Losing information is like losing an arm: employee reactions to data loss” (2019). Tamar is also a representative of the Israeli colleges on the University Libraries’ Research Support Committee.

Tamar carried out a small-scale study in order to gain a better understanding of the tools that researchers use to collaborate around data, and to explore the barriers and difficulties that prevent researchers from using institutional tools and services. Six semi-structured interviews were conducted with researchers from the University of Edinburgh, representing different schools, and all of whom collaborate with other researchers on a regular basis on either small- or large-scale projects. She found that participants use many different tools, both institutional and commercial, to collaborate, share, analyse and transfer documents and data files. Decisions about which tools to use are based on data types, data size, usability, network effect and whether their collaborators are in the same institution and country. Researchers tend to use institutional tools only if they are very simple and user friendly, if there is a special requirement for this from funders or principal investigators (PIs), or if it is directly beneficial for them from a data analysis perspective; sharing beyond the immediate collaboration is only a secondary concern. Researchers are generally well aware of the need to keep their data where it will be safe and backed-up, and are not concerned about the risk of data loss. A major issue was the need for tools that answer projects’ particular needs, therefore customisability and scope for interlinking with other systems is very important.

We’d like to thank Tamar for the great work she did, and for the beautiful olive oil and pistachios that she brought with her! Tamar’s findings will key into our ongoing plans for the next phase of the Research Data Service’s continual development, helping us assist researchers to share and work on their data collaboratively, within and beyond the University’s walls.

Martin Donnelly
Research Data Support Manager
Library and University Collections

Share