Jisc Learning Analytics: Defining learning analytics

Learning analytics has been defined by many
people. There are a number of different definitions
for it. One of the key ones is from a LAK conference,
that’s the Learning Analytics and Knowledge Conference, but what all of them really come
down to is the use of data about students and about the context in which they’re operating
in order to help those students and improve educational processes. What learning analytics isn’t is the kind
of institutional analytics that takes place, things about student finance, about predicting
student numbers. It really does home in on the learning aspects.
How to enhance things for individual learners as well as groups of learners and improve
all those processes around learning. Some people get learning analytics and learning
metrics confused. There is a difference between these things,
but really the starting point is data. So you have data from a number of different
sources. The two most commonly used data sources in
learning analytics are from the virtual learning environment and the student records system. But there are other systems, such as library
systems; attendance monitoring systems. And all these systems produce data from which
metrics are derived. So those metrics might be things such as the
number of times students attend lectures, how often they’re on campus, how often they
go into the virtual learning environment, the number of library books they borrow. Those metrics, in turn, can produce what’s
called composite metrics, which might, for example, be an indicator of the student’s
overall engagement with their studies. So you might combine the number of times they
logged into the virtual learning environment with their attendance on campus and even their
assessment, and produce this overall metric of engagement. That whole process, combined with the interpretation
of those metrics and the interventions that you might take with the individual students
to try and help them, for example, if they’re at risk of dropout or poor results. That is the process of learning analytics. So what’s new here? We’ve been looking at student engagement with
learning; trying to understand how students learn for decades. Learning analytics provides us with a whole
lot of new data sources and a more in-depth opportunity to analyse what students are doing. So never before have we had the ability to
home in on a student’s individual activity in the way that we can do now. And that’s partly because we’re doing a lot
more online learning and there are just more data points we can extract that student data
from. What traditional higher education has done
has provided a one-size-fits-all kind of educational experience for the students. So you deliver your course and you deliver
the same thing; the same lectures to students; you give them the same kind of assessments. With learning analytics we have the opportunity
to do a far more personalised approach here and provide individual help to students; pinpoint
those ones that are at risk and then take interventions to try and help those students. One of the strengths of learning analytics
is that we can use it at different levels throughout the institution. So we can start with the data about an individual
student, we can play that back to the student to show whether there are any areas that they
could improve in, for example. And the tutor can take that data as well and
provide more personalised input to the student. But that tutor or a lecturer might be responsible
for a wider cohort of students and can look at all the different students’ data together
comparing individuals with the average, for example. And then you might have a faculty structure
where a dean wants to see how the learners are progressing or even at an institutional
level. And finally, you might even want to export
the data outside the institution to government, for example, if the government wants to get
an idea, education department wants to see how students are doing across the board. We have to be careful here to be sure that
we’re not deceiving ourselves into thinking that learning analytics can explain everything
that’s going on. One of the premises of learning analytics
is that engagement is a proxy for learning. Of course, we can’t really know what’s going
on in the heads of individual students. We can measure it through assessments and
we can look at the amount of time they’re engaging with things. We might, for example, measure how often a
student is going into the library, how long they’re spending in the library, but they
might just be sitting in there talking to friends. So this is an important thing to remember
that causality is not the same as correlation. We’re correlating the amount of effort that
a student puts in with the final grade, but that doesn’t explain the whole picture. However, we do know that the most engaged
students tend to do better in the end. And that’s what learning analytics is trying
to tell us.

1 Comment

  1. Jago Brown

    March 12, 2019 at 2:37 pm

    Useful , simple , clear definition

Leave a Reply