By Linda Corrin, Sakinah Alhadad, Hazel Jones and Cassandra Colvin
As the field of learning analytics matures we are starting to see a shift in the way researchers and institutions are talking about learning analytics in the context of learning and teaching. From its roots in using student data to address retention, the field is expanding and becoming truly transdisciplinary – drawing on methodologies and analytics techniques from multiple disciplines. This provides exciting possibilities for new ways that learning analytics can be utilised in higher education.
In March 2018, the international Learning Analytics and Knowledge Conference (LAK18) was held in Australia for the first time. The conference, hosted by the University of Sydney attracted approximately 400 participants from across the world. The ASCILITE Learning Analytics SIG organisers attended the conference and observed the following trends that offer some insight into the future directions of research and development in the field.
An expansion of methodologies for learning analytics
Whilst traditionally there has been an emphasis on the use of quantitative methods for research in the learning analytics community, a shift towards a broader range of methodological approaches was evident across many parts of the conference program. This issue was placed front and centre as, in the opening keynote speech, Professor David Williamson Shaffer (University of Wisconsin-Madison) explored how the use of mixed methods could enable greater understanding of big data in education (David also gave a webinar on a similar topic for the ASCILITE LA-SIG in 2017 which can be accessed here). This perspective was in line with several pre-conference workshops that explored how both established and new methodological approaches could be applied to learning analytics. In addition, there continues to be a strong focus in the field on the use of multi-modal analytics which bring together many different data sources to help understand learning processes in the classroom and/or online.
The importance of design
Another key theme across many of the presentations and workshops was the importance of design in helping teachers and students to make meaning of learning analytics. Learning design was highlighted in several papers as central to the process of translating learning analytics into action. The best full paper focussed on the link between learning design, time on task and academic performance to be able to give more meaningful feedback to teachers and students (Nguyen, Huptych & Rienties, 2018). User-centred design, and more specifically user-adaptive visualisation, was the topic of the second keynote by Professor Cristina Conati (University of British Columbia). This work showcased the building of predictive models based on eye-tracking data in order to enable adaptive visualisations of data presented to users based on their relevant characteristics and needs. It was also pleasing to see more discussion about the role of students in the design of learning analytics innovations. A highlight of the multi-discipline nature of the conference was that the discussions of these ideas post-presentation was really vibrant – bringing together many different perspectives – challenging and stimulating thinking – and making participants reflect on how these ideas can work alongside those of their own discipline.
Greater integration of learning analytics with existing learning and teaching practices
Throughout the conference it was apparent that there is now greater integration of learning analytics within learning and teaching practices in higher education. Learning analytics is no longer seen as something separate. Instead data-informed approaches to supporting student learning are becoming common across a range of systems and activities. This is useful in ensuring that analytics is held to the same quality standards as other everyday learning and teaching practices.
Greater focus on the “learning” in learning analytics
It was evident that there is a continued shift in how researchers are considering “learning” within learning analytics research. Current research is moving beyond simply describing learning behaviour towards more sophisticated consideration of issues such as context and self-regulation. Ontological and epistemological framings of learning analytics research as a vehicle to understand learning were also brought under scrutiny by Professor Neil Selwyn’s (from Monash University) provocative keynote address on the final day of the conference.
Professional development for learning analytics
As learning analytics are being increasingly implemented in educational institutions, the topic of professional development of staff is emerging as a key challenge to be addressed. The best practitioner paper reported on an innovative model of staff development at Indiana University in which academics were encouraged to be leaders and innovators as learning analytics fellows. Discussion throughout the conference acknowledged the importance of the development of professional development opportunities for all staff who play a role in learning analytics in an institution.
Overall, the conference demonstrated that the future of learning analytics is bright and that institutions should be encouraged to take a strategic, holistic and transdisciplinary approach to its implementation.
The ASCILITE Learning Analytics SIG aims to facilitate conversations amongst ASCILITE members and the broader community about all these issues and more. In 2018 we will be continuing our webinar series with presentations by researchers from around the world on key learning analytics topics. We also encourage the sharing of ideas, questions and conversation in our LA-SIG Google Group or you can connect with LA-SIG Committee via email@example.com.