Many institutions now begin analysing student data from their VLEs. John Fritz gave an insight into what University of Maryland (UMBC) are doing with Blackboard and how they play back usage information to students in their CMA (check my activity) tool. London City University do similar things with their Moodle environment, and also my own institution, the Open University of the Netherlands, looks into Learning Analytics with our Moodle, Blackboard, and OpenU environments.

One thing that’s bugging the full utilisation of Learning Analytics, is privacy of student data. I am not really sure where these concerns originate from, and mostly it seems that they are made up on the fly rather than being based on solid legalistic arguments. But the effect is that, frequently, a group perspective of Learning Analytics is adopted to avoid the problem zone. This allows to anonymise a learner within a cohort, and results in emergent group patterns, such as how many percent of the students were active in chat.

This group perspective might be useful, for example, in evaluating functionalities or content items. It can highlight documents or dicsussion topics that are popular or unpopular with students. It can also show the course owner or tutor the most common learning paths through the online parts of a course.

However, in pedagogic terms and with respect to personalised learning support, such information is of limited value in helping learners to improve on their learning. Playing back the usage information to learners for reflection is of course another option, but in our survey earlier this autumn, the clear majority of expert respondents indicated that learners would require support to make the most of the Learning Analytics results. After all, what does it tell the learner that they have used the forum less than other students or less than last week?

This leaves the individual analytics that goes to the teacher unanonymised! The question is how this can really be a privacy issue, since there is already an established trust relationship and learning contract between these two parties, and it must be clear to everyone that evaluating engagement is part of a teacher’s job. In my mind, this is the best way how Learning Analytics can be usefully exploited for learning support.

Privacy is a real concern that shan’t be belittled as this news item shows, but there is no need for paranoia where established (teaching) processes and accepted trust relationships already exist. However, it maybe wise to create awareness and provide an opt-out clause for the unwilling.