Distance Learning

What are learning analytics

Part of educational innovation is closely linked to technological adoption. Digital whiteboards, mobile devices or virtual learning environments are witnesses and facilitators of the classroom revolution in terms of content and form. As we become aware of the challenges of the virtual world, the tip of a new iceberg approaches.

This time it is no longer through the use of one digital tool or another, but it is through the conjunction of all of them that a new paradigm is nourished: the analytics of learning. The term was born in 2011 by George Siemens and has to do with the collection and analysis of data from students in the same study environment. They have come to stay and we cannot ignore them in the future.

This new stage opens the way with a quantitative and measurable vision of education, with the first seed linked to MOOCs, mass courses and open distance learning. Virtual classrooms with thousands of students that offer golden opportunities to exploit this new oil that are personal data.

It is not surprising that in recent years multi-million dollar investments in educational technologies (the so-called EdTech) have been raining down. The promise is to improve the students’ experience, moving towards personalized and tailor-made learning. The foundation is the traces, the data, the fingerprint that students (and also teachers!) leave in the teaching-learning process.

These digital environments and technologies stand out for their ubiquity and the ability to capture multiple parameters in real time. All this information fills enormous and gourmand databases on online behaviour, performance, progress and student difficulties, to cite just a few examples.

But let’s take a step by step approach: no one will find it revolutionary to have a register with information on each and every one of the students: what are evaluations, reports and faculty meetings other than that? A series of indicators that are collected on each student that at the end of the period allow us to decide the grade, progression and needs.

Where is the difference? In scale and scope: so far the source of information is the eyes of teachers, professors and educators during school hours (in the classroom, the courtyard or the dining room). In the era of Big Data, virtual campuses, electronic books and mobile apps are added.

We move from the analog model of the notebook to ubiquitous spaces of automatic, detailed and permanent registration. An interesting and unprecedented range of opportunities opens up, but the coin always has two sides, and in issues of personal data and digital identity we are just beginning to see the importance of understanding the rules of the new game.

The new technologies are practical, simple and with infinite possibilities: complementing face-to-face work as a repository of resources, a discussion agora or an examination room.

And almost unintentionally, at the click of a mouse, it is noted how many times they enter the campus and for how long they stay, what pages they visit, what documents they download, what activities they deliver, how often they participate and how long it takes them to answer each question in the test.

Thus, each teacher can have automatically and at any time complete reports with the history of a student (or of the entire class), with the level of detail needed to assess whether there are too many suspended, too excellent and if the average grade is what I expected.

Assuming that there is a capacity to store and process this information (which requires specific infrastructure and knowledge), learning analytics can create value insofar as they facilitate the detection of anomalous situations, which move away from the “normal” pattern.

Let us imagine that we are interested in activating an alert for those students who have little chance of passing the course. School failure and absenteeism are a priority for David Pinyol and Miguel Ángel Carreras, members of the Learning Analytics group at the Open Institute of Catalonia. They are interested in detecting it in time to intervene before the student drops out.

The answer can be adapted, in each case, according to the history of information that is available for that student, but also learning from the records of previous experiences. It seems that the more data collected, the better predictions can be made. But this will always depend on the quality of the data, the reliability and accuracy of the indicators.

Experts agree that it marks a new look “beyond ideologies. It allows decisions to be made with the information in real-time”, the group in charge of applying and doing research on learning analytics at the UOC.

All these registers are also expected to be the basis of decision-making for the class, the cycle, the centre or the territorial administration. Along these lines, the UPC and the Generalitat promoted the Ágora project, which combines several resources for centres in Catalonia. So far there has been little experience, but the results do not always confirm the projections.