On the 20th of November I organized a master class at the well-organized Media and Learning Conference in Brussels on Learning Analytics and Visualisation of Data. I was grateful for this subject as visualisation is a very important aspect in the use of analytics. As a designer by background I know that sheets full of data will not influence the average user much. He neither has time nor the tendency to dive into piles of numbers the way some researchers do. But when presented timely and in a user-friendly and appealing way, data can make you change behavior.
I experienced this myself using sleep and movement tracking apps for a number of years. These so-called quantified self technologies are used in sports helping you to improve performance. Runkeeper and Strava are popular examples of such apps. I have used the Sleep Cycle app for over two years and discovered that I have the average deep sleep cycles of 1,5 hours. Also I found out that I have a certain specific amount of whole and quarters of hours of sleep I need to wake up refreshed. If my alarm clock wakes me up 15 minutes later I would again be in another cycle and feel more sleepy. So, for me to sleep longer was not always better. I discovered this easily by using the graphs that sleep cycle provided.
Later I used my Jawbone Up for tracking sleep, but mainly used this wearable as an activity tracker. It shocked me how little was the number of steps on an average office day. So instead of taking the tram to the train station I decided to walk. It only took me 5 minutes extra (which I won before by hacking my sleep ;-)). Also here the visible and direct feedback based on the data made me change my behavior.
It does not require much imagination to see opportunities for making data visible for analysis in the education context. During the master class I showed some examples of learning analytics tools using clear visualisations and also new ways of visualising data that could enhance learning analytics. You can view them in the slides below.
After the presentation we took the second of the two approaches of Duval and Verbert for Learning Analytics (Educational Data Mining (1) and Information Visualisation (2)) as a basis for a workshop focusing on the following questions:
- What are the needs of the learner/teacher?
How can this be visualized?
- What data is needed?
Focusing on the user’s needs offered a natural perspective for thinking of uses for learning analytics, of which some results are shown below. Unfortunately, time was too short to work out the questions to the full, but the process ignited the attendees to go on further independently.