Did you ever experience that, while teaching, you were able to point at students that will pass your course or in the contrary drop-out? Did you base these “predictions” on data, or was it your teachers instinct? Most likely you have developed a complex model to predict the success of your students, all based on certain variables and factors. But, what are these “factors” and how accurately do they help you to predict student performance? Although you might not been aware of this yet, this is the promise of the nowadays much discussed term learning analytics.
Learning Analytics can become a very important tool to continuously monitor and improve the level of the program.
To outline the basic definition, learning analytics are “the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimizing learning and the environments in which it occurs.” (Siemens, 2011)
Arising from that, management uses Learning Analytics as a tool to predict the success and retention of students. However, is that the only possible use? We believe that Learning Analytics should also be used to support teaching and improve the guidance of students in nearly real-time. This means:
- Bringing the insights from Learning Analytics to the lecturers and students.
- Making it real time.
For example, if a student lacks understanding on a certain topic, this can be seen in his online behavior. Using Learning Analytics the lecturer can be alerted immediately that the student’s online behavior has changed. This way the teacher can contact the student and they can work together on a solution.
Following both paths of prediction and description results in being able to predict the future, understand the past, but most importantly - asses the presence.
Using Learning analytics this way, it can become a very important tool to continuously monitor and improve the level of the program. When for instance the attendance rate drops, this can be noticed straight away and an evaluation can take place. This way we shorten the feedback loop drastically. So instead of evaluating once a term we can evaluate when it is necessary.
To summarize our thoughts, professor May (May, 2011) claims that learning analytics should be both descriptive, as well as predictive. In the case of predictive learning analytics, we want answers on questions such as: “Will the student pass his exam?” or “Does this student stay enrolled?”. When it comes to the descriptive perspective, learning analytics can be used for questions like: “Did the students understand today's class?” or “Is the student struggling?”, and “What actions are needed to help the student succeed?”.
Only by using Learning Analytics both for prediction and description the full potential can be reached. This can can be done by monitoring students in a real time and bringing the insights to the students and lecturers continuously. Following both paths of prediction and description results in being able to predict the future, understand the past, but most importantly - asses the presence.