At the BETT 2016 show Jan Hylén and myself shared some of our insights with implementing LA in a schools, on local level and national level. What have we learned that is important when doing implementations? We tackled issues like tool selection, privacy and privacy agreements, and a model/procedure to support the usage of data for decision making in a school.
The slides we used can be found here.
Concerning tool selection its important to look at whether a tool supports the vision of the school on personalised learning. Labeling a tool as supporting personalised learning does not tell you a whole lot. Not aligning the tools to the vision means tools can be very hard to work with or not being able to work with it at all.
A study looking at digital learning tools that all have some form of learning analytics functionality in the Netherlands looked at what kind of vision on personalised learning these tools supported. Most significantly is that these visions on learning changes the way the learning process is organised.
Teacher driven learning:
- Teacher directs the learning process of the group. Autonomous learning
- Learning materials are matched to the level of the group
- Learners are organized in groups depending on their age and level.
- Learning materials are aimed at the needs of the individual learner (pace, level, order).
- Learning is organized to work towards a pre-set goal.
- Teacher directs the learning process of the individual.
- Learner itself directs the learning process (with assistance and coaching)
- Learner is responsible for their own learning process.
- Learner learns on their own level, their own pace, in their own style, from their interests.
We found that most tools fit into the teacher-driven vision. Some tools fit into the autonomous learning model or would not require extensive modification to be able to support it. Self-organized learning is the least supported model. Which vision of personalised learning is supported can be seen most clearly from the way they organise information. Tools that support the classical model of organisation present their information mainly in the form of groups. Tools that are aimed more at gaining insights for an individual learner, regardless of any group they may be part off present information in a different manner. You can also see the vision reflected in the planning features of the tools which are aimed at groups (or subsets of groups within a larger group) or aimed at the individual.
We presented an example of a school in the Netherlands in which learners all have their own individual program and schedule, and which are not part of a pre-defined group based on age or level. Regarding tool selection they tell us the following:
“We have to look carefully at digital learning tools and whether they support our views on learning. Creators of digital learning tools often say that their tool supports personalized learning, but in the design of their tool they often work with the idea of groups or learners or classes, and not the individual. A lot of products therefor are not useful for us or can be hard to work with because they don’t fully support our way or working.”
An example on things you can do on a national level to help implement learning analytics that was presented was the privacy covent established in the Netherlands. A privacy covenant to protect personal data and provide safeguards for proper usage of personal data in education has been made developed in 2015. It contains a set of agreements on how personal data should be handled when using digital learning tools/materials and digital testing. The privacy covenant has been undersigned and developed in collaboration between school representatives and the major suppliers of digital learning tools/materials, student information systems, distributors of learning materials and ICT support suppliers.
A model contract has been also been made as a part of this covenant. A few points from the model contract were highlighted in the presentation. The agreements in this model contract will form the basis for agreements between schools and suppliers. The model contract describes what can and cannot be done with personal. It for example limits the use of personal data to the goals for which a school has given authorization. This limitation is in force during the running time of the contract and in the occasion that a contract expires or is terminated.
More information on the privacy covenant can be found on the Evidence Hub.
Model to support data driven decisions
Schools in Sweden and the Netherlands have worked with the model developed by researchers University of Twente to help making decisions with the usage of data. A school in Sweden for example looked at the question why their pupils were performing less well in Math than the pupils in the neighbouring school.
The model developed consists of eight steps:
- Problem definition
- Formulating hypotheses
- Data collection
- Data quality check
- Data analysis
- Interpretation and conclusions
- Implementing improvement measures
The University of Twente has developed a video explaining their methodology more in-depth.