Targeted analytics

I recently attended a talk by Professor Gregor Kennedy detailing his recent research into “learning analytics”, the study of interactions in digital learning environments. Professor Kennedy’s research has revealed the gap between curriculum design and real student behaviour, and the impact this has on learning.

We listened to details of Professor Kennedy’s recent research on interactions in digital learning (virtual learning environments: which to the University of Edinburgh means systems like Learn, Moodle and our MOOC platforms). In this research, he seeks to find actionable insight in the “footprint” of clicks, sequences and time within the massive amount of data contained in student interactions within online courses.

Professor Gregor Kennedy’s University of Melbourne profile

The key takeaways for me were whether we can explicitly say that behaviour (clicks, downloads, time) indicate ‘cognition’ – that is, has a student who read a group of pages, taking a certain time, actually learned anything about the topic?

And if, in the structure of these courses, there is no guide on how to conduct tasks, should we be surprised that students don’t do as ‘expected’?

Professor Kennedy went on to detail how learning analytics can inform this, and revealed exciting new research that aims to provide automated feedback (instant feedback created by systems to aid learning). An example given was of a system that provided automated feedback on attempts by surgical students on a middle ear procedure, without the intervention of an expert surgeon.

Developing Effective Automated Feedback in Temporal Bone Surgery Simulation

The data available from these courses can reveal learning ‘strategies’ (what pages were read, in what order, and how often); these strategies can reveal whether tasks are addressed as designed, or whether they are working well.

From the perspective of someone interested in web interactions: how can we use Web Analytics (and here, I mean Google Analytics, as that’s the system I use most often) to tell us about how our users have engaged with our sites? We should long ago have moved on from simple counting of hits on our pages, as that indicates nothing about understanding of the content of our pages. We should instead analyse smarter, and consider the order of pages read, the time on each page, and the interactions within them. Did web visitors read the pages we ‘expected’ them to, and in the order we laid them out? And looking at the ‘cognition’ of our (target) audiences, can we do more to design our sites to make it easier for our visitors to understand? On the web, we should not understate the importance of linking between pages that refer to each other, and in consistent ways: ‘calls to action’ at the foot of pages that give our visitors something to read next make it easy to know what information follows what we’ve just read.

The importance of structure and ‘calls to action’ – website style guide

Related links

Add comment: