Summary of activities 2019

If I was to try to define what a university is for – in the briefest of statements – then it would be the pursuit of knowledge. This easily covers research (the pursuit of new knowledge) and education (instructing others in the pursuit of knowledge). Therein lies the altruism within the university system where, as academics, we are interested in curiosity driven understanding in its purest form: because it is simply unknown to us. With the discovery of new knowledge comes an academic responsibility to share it (Smith, 2014). Why? Because, as I said in the 2014 Editorial, we should close the loop on understanding where it is subsequently applied to the real world and therefore gains social utility. In short, it benefits others. Sharing – of which peer-reviewed journal publication is considered the gold standard – is therefore often seen as an end-goal in and of itself. Yes, it benefits others, but because universities now use publications as a measure of academic success, it benefits the individual. As a result of this driver, there has subsequently been a dramatic rise in the volume of research published (Plume & van Weijen, 2014). This has also focused attention upon the funding of journals through subscriptions. Whilst ‘the reader pays’ is the most common model in any kind of publishing, it breaks the academic requirements for freedom of access to knowledge. If the model is ‘flipped’ then it becomes ‘author pays’ and so the Open Access (OA) journal is born. In truth there have always been journals that were freely available, however, their increasing prevalence and broad policy support by government makes them the preferred choice. OA journals were perhaps not inevitable, but their popularity has coincided with the adoption of topdown metrics by universities where all outputs are counted and bigger is better – more citations, more downloads, more likes, more shares. Just more. However, this has also been mirrored by bottom-up selfpromotion through social media. The central thread is OA – that is, the freedom to access academic material. Yes, there is the journal article, but also videos, live streams, in-person presentations, figures, workshops and, increasingly, data. Data touches upon different motivations in the academic, the university, the funder, and the reader. The collection, archiving, distribution, and downstream utilisation of data is one of the key benefits of research, sitting alongside new knowledge. Data sharing is common in some subjects, particularly where there are experimental results that need external verification and validation. What OA publishing has precipitated is a move toward the formal lodgement of data that are now seen as an integral part of the research outcomes and self-promotion. Academic publishers have long supported the lodgement of supplementary materials, including data, whilst Journal of Maps has supported data publication since our first issue. Uniquely, we embedded datasets within the PDF so that the data genuinely followed the article; however, this wasn’t a sustainable solution as it both breaks PDF/A compliance (Library of Congress, 2019) and is not scalable to large datasets. Ad hoc online publication and, latterly, data repositories have a much longer history although the extent of their use varies by subject. I’ve said that the key elements of OA should be to allow verification and validation. In summary: are my conclusions supported by the data collected and analysis undertaken and, if that is the case, can they be repeated independently? Verification, therefore, necessitates the key requirement of being reproducible. Not only should the data be available, but exactly the same analysis should be achievable to verify the results. The logical end-point is for the use of freely available software (ideally open source), a programmable environment, the original scripts and a versioning system, both for the environment and the scripts themselves. Within this context, the open source statistical programming language R (https://www.r-project.org) presents itself, alongside a versioning environment such as Git (https://git-scm.com/). Like most software, R releases new versions regularly; however, by using the containerisation environment of Docker (https:// www.docker.com/) you can run a script within exactly the same version of the R core and requisite plugins. Therefore, a journal article can describe what was undertaken, present the results and then enable the reader to verify that the outcomes are genuine, as