I published the last blog before I had completed it – which is why it seemed so unfinished… because it was!

The Society for Learning Analytics Research defined learning analytics as

“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”

(Long, P.D., Siemens, G., Conole, G. and Gašević, D.(Eds) (2011), Proceedings of the 1st International Conference On Learning Analytics And Knowledge (LAK’11), ACM, New York, NY)

Learning analytics are increasingly part of the landscape of education.

One of the main reasons is because, in recent years, we have become obsessed with a desire for a perceived greater efficiency in all walks of life. And education has been caught up in this societal drive. To be better ‘producers’ of education, institutions try to understand how learners function and how better they can provide a more efficient, stream-lined ‘service’. Institutions are expected to provide an educational provision that is seen to respond to the needs of the learners, and, in turn, learning analytics are seen as a pragmatic fix to expedite this provision. Biesta’s concept of learnification where learning becomes tied up with matters of accountability and productivity, where learning is reduced to good outcomes, meaning that it becomes a means-ends mechanism, has played into the hands of the development of learning analytics. We are in danger of reducing education to consumer learners and producer institutions (Gasevic et al.)

Something that really resonated with me is an idea posited by Gasevic et al., which is that because face-to-face learning is happening less, therefore we have fewer social cues and therefore the ability to pick up on a student’s engagement is less. In an increasingly online world, teachers are starting to rely on algorithms and data to give them information that previously they would have gleaned from multiple meetings and tutorials with their students. How else might you now judge the learning behaviour of your students?

But what can you learn from data? It is all about giving a window into behaviour. However, I would argue that that window is more about analysing what your learners do, rather than how they learn.  Learning analytics have a place for providing an indication of general behaviour, but are unable to be actionable or meaningful for an individual learner.

An example would be the ability to check, monitor and measure user frequency on a learning platform such as Moodle or Firefly. That is not really a quantifiable indication of learning. I could log in many times and thereby show huge engagement with a course, but that does not show I have learnt anything. Learning analytics can be a poor substitute for meaningful engagement and learning. Too often, they are rather more ‘activity analytics’, a phrase neatly coined by Wilson.

Are we now generating a blurry and convoluted datafication that is obscuring what matters?

One of the problems is that it is not the educationalists that are running the data analytics. As Wilson points out, institutions often use the same off-the-shelf data and algorithms to monitor learning, which does nothing to meaningfully measure individual learners and their individual learning needs.

This idea of off-the-shelf analytics can lead to a homogenization of education, whereby learners can be ‘nudged’ by recommender systems as to what they should read, discuss and watch. Much like Netflix’s recommendations. This leads to creativity and experimentation, two essentials of learning to be at the risk of becoming stifled (Greller and Drachsler).

Learning analytics have a place in a consumer-driven society. However, I don’t believe they are particularly useful at an individual level.

Learning Analytics #2 / Emma Morton / An introduction to digital environments for learning (2020/2021)[SEM2] by blogadmin is licensed under a Creative Commons Attribution CC BY 3.0