Archive for category Events
On 14th December, Duygu Bektik defended her thesis successfully, and now only minor corrections stand between her and her doctorate.
Learning Analytics for Academic Writing through Automatic Identification of Meta-Discourse
When assessing student writing, tutors look for ability to present well-reasoned arguments, signalled by elements of meta-discourse. Some natural language processing systems can detect rhetorical moves in scholarly texts, but no previous work has investigated whether these tools can analyse student writing reliably. Duygu’s thesis evaluates the Xerox Incremental Parser (XIP), sets out ways in which it could be changed to support the analysis of student writing and proposes how its output could be delivered to tutors. It also investigates how tutors define the quality of undergraduate writing and identifies key elements that can be used to identify good student writing in the social sciences.
On 13 December, I joined a Foresight Workshop on Learning Technologies in Luxembourg. The workshop was designed to help the European Commission to set and define future European strategic research and innovation priorities.
The workshop began with a series of ‘Moonshots’. Individual experts presented ambitious, yet realistic, targets for EU-funded learning technology research and innovation up to 2025. For each of these, we considered: What is the problem? How is it dealt with now? What difference would it make if this problem were addressed successfully?
We went on to merge our individual Moonshots into Constellations and then into Galaxies. We made links between the different ideas, linking them with other international activities and trends, as well as to previous EU-funded work. I was interested to see that many of the experts from across Europe presented ideas associated with blockchain for learning, a pedagogy that was picked up in our recent Innovating Pedagogy report.
My moonshot focused on a series of problems: access to tertiary education is unequal, most people in Europe do not complete tertiary education and many people in Europe need to develop new skills. Massive open online courses (MOOCs) offer a potential solution, but these new approaches to learning require new approaches to teaching. Teachers need training and support to work effectively in these new environments. They also need proven models of good practice. Improving educator effectiveness on these courses has the potential to increase Europe’s capacity to respond to its priority areas. It also has the potential to open up education for millions by developing and sharing knowledge of how to teach at scale.
The PELARS project (Practice-based Experiential Learning Analytics Research And Support) invited me to Brussels for their Policies for using Big Data event on 9 November. The aim of the workshop was to raise awareness about the potential of analysis of data produced by learning technologies to catalyze the effective design of adaptive teaching, learning and assessment at scale. The aim was to bring together people interested in exploring the state-of-the-art of learning analytics, as well as to be informed about opportunities and barriers for adoption.
I chaired the panel discussion at the event, and was also able to talk to participants about the LACE project, following a presentation on LACE by Hendrik Drachsler.
Il-Hyun talked about the problems associated with learning analytics in a country where grades are allocated in relation to a normal distribution curve – so if one student’s grades go up, another student’s grades will go do – and where competition to enter universities is so intense that retention is not viewed as a problem.
While I was in Seoul in September, I took part in the Asian Learning Analytics Summer Institute (LASI Asia). I was joined there by members of the LACE team, who included the event as part of the LACE tour of Asia, which also took in Japan and Korea.
During LASI Asia, I gave a talk about what is on the horizon for learning analytics. This went into more detail, and was aimed at a more specialist audience, than my talk at e-Learning Korea. I also took part in a couple of panel discussions. The first was on how to build an international community on learning analytics research, and the second was on the achievements of learning analytics research and next steps.
There is general agreement that the importance of learning analytics is likely to increase in the coming decade. However, little guidance for policy makers has been forthcoming from the technologists, educationalists and teachers who are driving the development of learning analytics. The Visions of the Future study was carried out by the LACE project in to order to provide some perspectives that could feed into the policy process.
The study took the form of a ‘policy Delphi’, which is to say that it was not concerned with certainty about the future, but rather focused on understanding the trends issues which will be driving the field forward in the coming years. The project partners developed eight visions of the future of learning analytics in 2025. These visions were shared with invited experts and LACE contacts through an online questionnaire, and consultation with stakeholders was carried out at events. Respondents were asked to rate the visions in terms of their feasibility and desirability, and the actions which should be taken in the light of their judgements. 487 responses to visions were received from 133 people. The views of the respondents on how the future may evolve are both interesting and entertaining. More significantly, analysis of the ratings and free text responses showed that for the experts and practitioners who engaged in the study, there was a consensus around a number of points which are shaping the future of learning analytics.
1. There is a lot of enthusiasm for Learning Analytics, but concern that its potential will not be fulfilled. It is therefore appropriate for policy makers to take a role.
2. Policies and infrastructure are necessary to strengthen the rights of the data subject.
3. Interoperability specifications and open infrastructures are an essential enabling technology. These can support the rights of the data subject, and ensure control of analytics processes at the appropriate level.
4. Learning analytics should not imply automation of teaching and learning.
The full results of the study are published in a report at http://www.laceproject.eu/deliverables/d3-2-visions-of-the-future-2/.
In this session the visions explored by the LACE study will be presented, the conclusions discussed, and the audience will take part in an impromptu mapping of the most desirable and feasible vision of the future for learning analytics in Asia.
While I was in Montevideo, at the invitation of Plan Ceibal, I was interviewed about learning analytics. This playlist of four short videos (subtitled in Spanish) deals with the potential of Big Data to improve learning, how The Open University has used learning analytics, and the work of the LACE and LAEP projects.
I talk about how analytics can be used to identify when students are dropping behind, how they can be used to identify successful routes through courses, and how they can identify types of learning design that lead to student success.
I note that the supply of learning analytics is growing, but it is not clear that the demand is growing in the same way. Researchers and developers need to engage more with educators at every stage in order to identify the problems they need to be solved and the questions that they need to have answered.
I also talk about the need to align learning analytics with strategic priorities for education and training, not only at institutional level, but also at national and international level.
My videos are followed in the playlist with videos from Professor Dragan Gasevic, chair of the Society for Learning Analytics (SoLAR).
I visited the University of Deusto in Bilbao, Spain, to give a keynote at the learning analytics summer institute there (LASI Bilbao 2016) on 28 June 2016. The event brought people together from the Spanish Network of Learning Analytics (SNOLA), which was responsible for organising the event, in conjunction with the international Society for Learning Analytics Research (SoLAR).
What does the future hold for learning analytics? In terms of Europe’s priorities for learning and training, they will need to support relevant and high-quality knowledge, skills and competences developed throughout lifelong learning. More specifically, they should improve the quality and efficiency of education and training, enhance creativity and innovation, and focus on learning outcomes in areas such as employability, active-citizenship and well-being. This is a tall order and, in order to achieve it, we need to consider how our work fits into the larger picture. Drawing on the outcomes of two recent European studies, Rebecca will discuss how we can avoid potential pitfalls and develop an action plan that will drive the development of analytics that enhance both learning and teaching.