Assessment analytics see http://publications.cetis.ac.uk/2013/750 Much of the research in this area has been carried out, but a wider questions are still out standing, some of which are listed in projects already on this site. The benefit of focusing on this area is that it will provide greater benefit to 'blended' courses, and target data collection on maximising the type of data needed to make a deeper analysis of... more »
We need to move away from seeing Learning Analytics as being 'just' about diagnostics (spotting when a student is likely to fail so that we can make an intervention) and start to develop the use of analytics as a tool to support us in assessing students' competences, dispositions, and trajectory within their community of learning. This is about analytics as a tool to help us develop more authentic, 'real world' relevant... more »
Undertake desk research to assess the kinds of analytics being undertaken by MOOCs and to evaluate the viability of adopting the same approaches in more traditional educational contexts.
'Human readiness' for learning analytics isn't a given. It's important to ask under what circumstances learning analytics might be welcomed by teaching staff, students, and other decision-makers - in principle and practice. The work of Anna Lea Dyckhoff advances a view of learning analytics as answering action research questions posed by teaching staff. Carrying out similar studies for other stakeholders - particularly... more »
Develop a code of practice for institutions on use of learner analytics and student data. The code should explore ethical issues, good and bad practice, legal requirements and how to do learner analytics in a positive way.
Produce a student owned data dashboard that gathers any data related to the student, which may include learning attainment, engagement activity and personal data (e.g. sleep patterns, etc) and provides a view for them of how they are coping with learning/student life.
A network and guide to support approaches used by institutions to a) support learners identified as at risk b) to modify the institutional behaviour to improve learning attainment. This could be a blog/online resource and series of workshops/network meetings (which could be led by membership bodies?)
Investigate what roles institutions needs to support different learner analytics approaches (and what it would cost them).
Provide opportunities for institutions at similar stages of learner analytics implementation to work together and share what they are doing through a series of special interest groups. Suggested groups could be around course dashboards, attendance tracking, ethics, BME and gender groups, etc.
Building on the work of the Library Analytics and Metrics Project (LAMP). Develop further use cases to extend the data set and look at learner analytics in relation to resource usage.
Survey institutions to benchmark what they are already doing with learner analytics. This would benchmark the sector and also help to determine the maturity of institutions in relation to learner analytics i.e. data warehousing, tracking, traffic lighting, predictive analytics, personalised data
Undertake a short study (3 months) to review existing institutional and commercial learner analytics tools to allow institutions to make informed choices on the best solutions for them.
The most expensive resources for most institution are the estates and infrastructure. The approach would be to develop a set of tools that allow institutions to examine use of estates for learning (i.e. teaching rooms) to ensure efficient use of rooms; explore staff workloads to help manage and costs courses.
A report to inform institutions on what specific data they need to gather to do effective learner analytics. Combined with the metrics this will avoid all institutions from gather vast amounts of data unnecessarily and wasting time doing their own analytics to determine what correlations are significant for predicting student attainment.
Research existing learner analytics and produce a basic summary of the metrics (what data matters and how to analyse it). This study would also provide some key learning from existing learner analytics that could be used by institutions to make informed decisions about their choice of analytics tools and how to interpret information. The report would also support a more DIY solution to the course/student dashboard.