Learning Analytics – making learning better? The Dutch perspective


On the Friday afternoon of BETT 2015 I had the pleasure to take part in the panel of the LACE seminar on learning analytics. The seminar aimed at the schools sector was led by LACE partner Dr Jan Hylén, working with Skolverket in Sweden and started with an excellent introduction to learning analytics and educational data mining by LACE partner Dr Doug Clow from The Open University.

My short presentation concerned the viewpoint from the Dutch national schools network. For Kennisnet, the public educational organization which supports and inspires Dutch primary, secondary and vocational institutions in the effective use of ict, learning analytics is a key topic. Many schools in the Netherlands are working on the implementation of personalised learning and technology like learning analytics always plays an important role.

Data-driven Education with embedded and extracted analytics

Because of its impact we dedicated an entire chapter on data, learning analytics and its application in adaptive learning materials and personalised learning environments in our 2014-2015 Trend report ‘Technology compass for education’ provocatively called ‘Data-driven education’. Obviously we don’t want data to ‘drive’ learning, but merely to enable improving it’s quality and flexibility.
As Dutch schools are rapidly adopting digital learning materials, cloud-based learning management systems and many other small digital tools and apps, all learning activity leaves digital data traces. These data traces offer information on learning as it happens and once collected and analysed they can be applied to adapt the learning process of the individual student.
Many researchers make a very useful distinction in two types of learning analytics: 1) embedded and 2) extracted.

  1. Embedded analytics refers to the use of data from the learning proces to alter the experience dynamically and automatically, as it happens, without human intervention. Learning materials that use data this way are called adaptive and are often compared with games which apply similar strategies to keep their players engaged.
  2. Extracted analytics refers to the use of data to reflect on the learning process and to inform/advise teachers (among others) on interventions ranging from adjustments in individual learning processes to evaluating and changing national curricula.

The infographic below shows data as the ‘raw material’, processed by learning analytics and applied in embedded and extracted applications. Both cycles feed themselves as new data from the adjusted process enables new adjustments.

Learning Analytics: Embedded and Extracted

Learning Analytics: Embedded and Extracted

Application of learning analytics in Dutch schools

Especially in primay education adaptive learning materials on math and (Dutch) language skills have been around for some years and are used in many schools, although almost always as an addition to traditional learning materials. The Dutch company oefenweb.nl for example offers Math Garden (used in 10% of Dutch primay schools) and Language Sea. These learning materials are widely used by schools implementing tablets because they offer each student the opportunity to work at their own pace as the software adjusts the level of difficulty of the problems that are offered. At the same time the teacher has a ‘back-office’ which shows the progress of each individual student on the different topics covered by the material. Examples of schools which heavily use tablets are the so-called SteveJobs schools or O4NT (Education for a new Era). Other suppliers supporting tablet education are snappet which focusses on supporting the teacher when differentiating between groups of students needing different support and challenges.

The availability and use of personalised learning environments supporting dynamic learning processes/routes for students is still very limited. One interesting example of a platform that uses data analysis to offer a personalised learning solution is PulseOn. This technology is used by several small groups of schools (i.e. Learntoo) doing limited scale experiments. Dutch publisher Malmberg (part of Sanoma learning) has started an interesting partnership with US based adaptive learning solutions company Knewton which will result in adaptive learning courses later in 2015.

What about Privacy?

Using detailed information on the learning activities of children naturally results in tough questions about the protection of their privacy. There is is a justifiable public concern for the privacy of both students and their teachers. Who should be able to access which data, and when and where, and for what purpose? Many of the questions asked during the seminar were related to this issue.

As the Dutch Parliament has expressed its concerns about the collection of student data by publishers, Kennisnet assists Dutch schools to implement policies that respect Dutch law on the protection of personal data. At the same time Kennisnet is exploring effective frameworks like Privacy by Design and a User Centric approach that can address the ethical challenges that arise from the use of the data traces generated by student learning, looking for patterns and behaviors. Privacy by Design advocates that designers of learning materials ensure privacy in an application from the earliest stages of design. The User Centric approach aims to put the user in control of their own data and its uses. Clearly we have only started to adress these very important issues. Currently the European partners are exploring the possibility to collaborate on resolving these issues.

Kennisnet and LACE in 2015

The European Commission supported LACE project has highlighted and presented interesting research on the use of learning analytics in 2014. The team at Kennisnet is looking forward to continue working with our international colleagues to share new insights with the international education community.

View Twitter conversations and metrics using [Topsy]


About Author

Michael is a Strategic Advisor on Innovation at Kennisnet, the public educational organization which supports and inspires Dutch primary, secondary and vocational institutions in the effective use of new technology. Michael is a computer scientist by background and is specialized in helping schools understand and apply new technology to support more engaging and more effective learning. Michael focuses mainly on two areas: 1) advising schools about future-proof, reliable ICT infrastructure that fits their specific needs and 2) stimulating awareness among schools on the potential of data and learning analytics to facilitate data-enabled education while respecting privacy of both learners and teachers. His many ‘hobby-projects’ include the promotion of teaching computer science and programming in schools and the internet of things including but not limited to the quantified self.

Leave A Reply