Research Data MANTRA: A Labour of Love

Research Data MANTRA (or Management Training) is a labour of love. It has been an integral part of the University of Edinburgh's Research Data Management (RDM) programme since 2012. The staff at EDINA and Data Library at the University of Edinburgh has been curating this resource, based on internal and external feedback and has just published its fourth release since 2011. MANTRA is an open, web-based training course intended for self-paced learning by PhD students and early career researchers or others who manage digital data as part of a research project. It informs about good practice in research data management with real life stories. MANTRA’s approach is to be fun, relevant, useful, interactive and timely (FRUIT!). Librarians’ training needs are catered for through a companion resource, the DIY RDM Training Kit for Librarians.


Introduction
The Research Data MANTRA (Management Training) online course is an integral part of the University of Edinburgh's Research Data Management (RDM) programme delivered by Information Services.Since it was created in 2011, the University of Edinburgh's Institute for Academic Development (IAD) has helped us introduce it to PhD students and early career researchers who have valued the course.The website is managed by EDINA and Data Library, a division of Information Services, with both national and local remit, extending into UK higher and further education and beyond.Thanks to a continuous positive feedback loop from users and web logs, this resource has avoided the fate of some completed project outputs-slow decay and eventual abandonment.To the contrary, the Data Library team continues to nurture and develop content and sustain maintenance.Now that Edinburgh University is pursuing a data-driven research agenda through the advent of Edinburgh Data Science (Bayne, et al. 2014), it is more important than ever that we support our students and researchers to attain good practice in their data analysis and data management.Because MAN-TRA is openly licensed and freely available, there is no limit to who may benefit.Our usage statistics are limited due to the open nature of the course (no registration, no assessment), but Google Analytics has recorded nearly 10,000 new sessions from people in 144 countries over the last academic year, and 16,000 total sessions (Rice, 2014).

What is MANTRA?
The Research Data MANTRA course is a free online training course intended for selfpaced learning by PhD students and early career researchers or others planning to manage digital data as part of the research process.It is available online at http:// datalib.edina.ac.uk/mantra or for download as an open educational resource (OER) from the Jorum repository, http://find.jorum.ac.uk/ resources/18281.It was developed at EDI-NA and Data Library, University of Edinburgh, by two data librarians and an elearning specialist, with funding from Jisc in 2010-11, under the Managing Research Data Programme's training strand (Jisc, 2010).
The course aims to increase researchers' awareness of data management issues.It informs about good practice in research with real life stories and scenarios.The eight online learning modules take an hour or less to complete, and are designed to be fun, relevant, useful, interactive, and timely (FRUIT).For example, images are used to break up text, short video clips of researchers talking about their own experience appear in each unit (Figure 1), factual scenarios encountered by researchers in different disciplines are conveyed, and interactive quizzes about the content is presented at regular intervals.
Accompanying PDF tutorials give a deeper grounding in data handling skills, such as importing and exporting data, recoding, and keeping track of transformations within four common software packages using open, empirical datasets, available for download from the website.The analysis environments covered are NVivo, SPSS, ArcGIS, and R  Data protection, rights and access: ethical obligations towards human subjects, it is staff that faces the greatest culture change with regard to RDM requirements and expectations.
Although we were funded to create discipline -specific training materials, we really wanted to try to create something compelling across a broad range of disciplines (or 'schools' at our University-the equivalent of academic department).So we targeted three schools with post-graduate programmes, and created one resource for all three, in the hope that they would be representative enough to work for others as well.The programmes were Social and Political Studies (SPS)which had both quantitative and qualitative data analysis requirements: the Doctorate in Clinical Psychology-which was unusual in that much of the programme was offered as distance education, and students spent much time in the field with National Health Service staff supervisors; and GeoSciences-with whom, as with SPS, we had much prior contact.The three schools happily spanned the overarching colleges of the University.All shared a concern with managing confidential personal data.We felt-and still do, confirmed by anecdotal feedback-that we had at least covered the ground of disciplines where there is a 'lone researcher,' if not research groups using sophisticated computing systems.
Before writing the learning materials, we conducted a needs assessment with PhD training managers in each school based on a single recorded face-to-face interview with each.The IAD director had advised us not to expect too much time from these project stakeholders, so we were careful to try to get what we needed from one interview (Rice, 2010).Over the one-year project, schools experienced a high amount of academic staff turnover-maternity leave, sabbaticals, promotions and exits-making it difficult for us to meet their successors and encourage follow-through on their end.Each school managed to 'embed' MANTRA into its training curriculum either through induction and training activities or in the required coursework, trying to create culture change-with or without the big sticks of the funders' mandates.
If we talk about good practice in data management often and loud enough, it may happen (like a mantra).
An academic at our institution once challenged our use of the colourful image on our flyer (http://edin.ac/1pgpTop),seen partially on the home page banner-of a Tibetan monk's hands carefully adding the last grains of sand to a beautiful sand mandala.She said the purpose of creating a sand mandala is so it can be destroyed, symbolising lack of attachment-not exactly a symbol of sustainability.Her point was well-taken, but we felt that our use of this image can symbolise how easily one's data can be lost if precautions are not taken.It has happened to everyone on a small or grand scale.Moreover, the hands of the monk can symbolise the actions of the data curator to care for the dataset.

The Origins of MANTRA
Originally, we were funded to produce training for PhD students: "JISC Call for Projects 04/10 sought to fund projects to encourage research data management training to be embedded in post-graduate academic curricula" (Jisc, 2013).We had already done a pilot RDM training course with some GeoSciences doctoral students in December, 2009 and felt we could apply the experience (both positive and negative outcomes) to a larger project.
Our project partner, the Institute for Academic Development, convinced us to a) run the course as a stand-alone online course using the model of PG Essentials, purchased from Melbourne University containing optional, self-paced online modules such as 'writing a literature review' that appeal to new PhD students for their pragmatism; and b) to include early career researchers (aka post-docs), as they too are working on their research skills and are within the target audience of IAD.This seemed entirely reasonable to us, since although it has remained an optional activity, even in required courses.In the GeoSciences, we have been invited by one of the successors into the classroom to introduce MANTRA and help the students draft a Data Management Plan (DMP) for their research proposals.However, we have not been involved in evaluating the DMPs.
The school-based stakeholders also helped us identify the most useful analysis environments on which to focus the data handling tutorials for them and agree to the layout and learning objectives with each of the commissioned authors (experts in each software package).The idea of the data handling tutorials came from the Data Library and was included in the original funding bid as a 'unique selling point.'Our experience with the pilot workshop indicated students wanted hands-on exercises as well as being told general principles.As eager new researchers, they tend to be more interested in data analysis than data management, and our experience as data librarians told us that data handling could be a bridging skill facilitating both.For example, to conduct statistical analysis on a large dataset, certain preparation is needed, especially on other people's data.Moreover, the Data Library led an earlier project aimed at bringing 'real data' into learning environments to raise students' quantitative skills, particularly in the social sciences, so this was a way to further that aim as well (Rice and Fairgrieve, 2003).
One finding that emerged from the needs assessment was that unlike ourselves, the doctoral trainers (at that time) were not concerned with teaching their students to share or preserve their data at all, never mind how to do this well.This may have been partly due to the timeframe (the funders' requirements that have driven awareness were still emerging) and partly due to the bias that a novice's data is not worth sharing or preserving.For this reason, the final two units of MANTRA were 'shelved' for a second release and were not finished during the course of the funded project.However, they were both released in a second 2012 version as we viewed the topics as crucial to the spirit of RDM.As with the first six units, we conducted user testing with doctoral students from the three schools to ensure the usability and usefulness of the new material.
An external evaluation of the MANTRA project was completed in 2012 by a staff member of the Digital Curation Centre based at the University of Glasgow, after the course had been available for one academic year (Molloy, 2012).By this time MANTRA had developed a good reputation in the emerging 'RDM community.'However, evidence was limited by our basic web statistics, in that we did not have the full tracking capability of user behaviour that a Virtual Learning Environment could have captured.(Google 'universal' analytics with event tracking may provide new insights over the coming year.)

Lessons Learned for the Third (2013) Release
The third, 2013 release combined minor cumulative edits to the online module, a major change in the presentation of the course via the home page -done to our specification by a web and graphic designer, and a back-end upgrade to the XOT software which resulted in html5-compliant web pages.The significance of moving to this standard and away from Flash meant that the units could be viewed on iPads for the first time: a significant user requirement for us.
Since our University's RDM Policy was passed ( 2011) and an RDM Roadmap published ( 2012), the RDM Steering Group that governed the programme saw fit to test MANTRA across the University.However, they and their colleagues were mainly in the College of Sciences and Engineering, particularly Physics and Informatics, so there was some concern that the first unit, Research data explained, was too basic (the Big Data section had not yet been written).
We also had feedback from the Institute for in this kit to train all of our academic service librarians, co-facilitated by a data librarian and an academic service librarian.The idea is that with the help of a good facilitator (i.e., someone who is able to stimulate discussion and involve the entire group), the training can be done without an 'expert' instructor.In fact, the facilitator can be one of the learners, or the role could even be shared amongst the participants.
The training comprises five two-hour, face-to -face sessions, with learning and preparation time between each session of two weeks or more.The sessions open with short talks (from local speakers or alternatively, podcasts from the kit) followed by group exercises created by the UK Data Archive, with lots of supportive, small group discussion.MAN-TRA modules are used as reading assignments and accompanying 'reflective questions' are used to help librarians 'put themselves in the shoes of the researcher.' Learning is reinforced and put into practice through an independent study assignment of completing and publishing an interview with a researcher, resulting in a simple data curation profile-ideally, to be publicly shared.The independent study is optional, since the target group is busy professionals, though we feel it is crucial for building confidence needed to take the training into practice, and also to appreciate the anxiety many researchers feel about being 'forced' to share their own data online.
Academic Development (IAD) staff that learners need to know what they're going to get out of a resource before they invest their time in it.So the third, 2013 release of MANTRA included four personas in the top right corner (Figure 2).If a 'career researcher' clicks on their persona, the page suggests that they skip over the first unit and begin with Data Management Plans.'Research students' can start with the first unit, whereas 'senior academics' are shown the Data handling tutorials for their students, and to encourage culture change, the Sharing, preservation and licensing unit.We also added some JavaScript that gives a peek at each unit's content without leaving the home page and tried to make the data handling tutorials stand out by using a different background colour on the button.
We suggest that information professionals start with Data protection, rights and access, moving onto Data sharing and preservation, since those are likely to be the topics they need to know best to give staff and students RDM support and advice.But we also have a special offering for librarians, described below.

The DIY RDM Training Kit for Librarians
To help professional librarians grasp supporting RDM, the DIY

What Are the Topics Covered?
 File formats and transformation: open vs proprietary data file formats, compression, normalisation, and other kinds of data transformations; why such skills are useful in a research context. Documentation, metadata, citation: the importance of documenting data during a research project, re-using other people's data, purposes of metadata; the importance of proper data citation and reproducible research. Storage and security: issues It was time to do a semi-major overhaul of the written content of the online modules, in order to keep MANTRA from fading into obsolescence.Data Library team members have edited additional video material, improved the language and flow of the units, suggested new content and references, and helped test, revamp, and update the data handling practical exercises (our small team has experienced some flux as well as growth thanks to the University's RDM programme).A change control document was used as we updated the online modules, partly to give Deliverables.Last updated July 31, 2013.http://www.ed.ac.uk/schools-departments/ information-services/about/organisation/edl/ data-library-projects/mantra/deliverables Rice, Robin.2014."New Release of Research Data MANTRA (Management Training) Online Course."Edinburgh Research Data Blog, September 5. http:// datablog.is.ed.ac.uk/2014/09/05/new-release -of-research-data-mantra-managementtraining-online-course/ Rice, Robin, and Joan Fairgrieve.2003."An Enquiry into the Use of Numeric Data in Learning and Teaching in UKHE."Radical Statistics.81:18-33.London: Radical Statistics.http://www.radstats.org.uk/no081/ rice.pdfDisclosure: The author reports no conflicts of interest.content in Journal of eScience Librarianship, unless otherwise noted, is licensed under a Creative Commons Attribution 4.0 International License.http://creativecommons.org/licenses/by/4.0ISSN 2161-3974 and Minho University has expressed interest in translating it into Portuguese.Anecdotally, librarians have expressed gratitude that it has saved them from developing their own researcher or librarian training from scratch.It has been a labour of love because the team has remained dedicated to its quality assurance, and to keep the course openly licensed and free to all comers.