An Expanding Vision for First-Generation Learning Analytics

Analytics in general is the processing of large amounts of data to help guide decision-making. Learning analytics applies that model to the goal of improving how well students learn in educational environments.

Most learning analytics systems center on online learning scenarios, where software can collect the data automatically. Using activities that are part of the online learning process — like logins, assignment completion, quiz scores and so on — learning analytics seeks to analyze the “digital breadcrumbs” that students leave as they interact with the online learning environment (and potentially other online spaces, too, such as Facebook).

Over time, as data piles up, a learning analytics system can gradually identify correlations between activities, the outcomes of activities, and learning outcomes. The type of data gathered, in general, includes information about how frequently students access online materials, and/or the results of assignments, tests, exercises and other online activities.

For example, learning analytics software could compare Student A’s activity with current classmates, students who took the same course last year, and students who took related courses in prior years, in an effort to predict how well Student A will do in the class. If Student A’s predicted outcome was non-optimal, the system would then initiate some action (such as alerting the instructor) in an effort to improve that outcome.

Here’s a synopsis of the best scenario/example I’ve seen on how learning analytics can help an educator guide students in an online class:

Picture an online “psych 101” class at a university, where a learning analytics program is tracking students’ digital activities: how often they login, what resources they download, how often they post to the course blog or discussion board, grades on assignments and tests, peer evaluations and attendance.

Each week, the learning analytics system analyzes the data it’s collected, searching for correlations between activities and learning outcomes. As it identifies learners whose activities suggest a low likelihood of success in the course, it automatically notifies them and makes suggestions for actions they could take to improve their odds.

Some students shrug off the automated messages, but some apply the guidance. The instructor is copied on the messages, and chooses to engage with some students as well. Many in this latter group respond to the extra support by increasing their level of participation.

Similarly, for students who are participating at a high level and are predicted to do well, the system provides reassurance that they’re likely to succeed.

The learning analytics system can also help guide the instructor to make changes. For example, if engagement among students drops during a specific unit (e.g., on a topic that students aren’t comfortable discussing openly), the instructor can develop a new set of “anonymous” exercises to encourage more participation.

Learning analytics tools can track far more data and activity, far more consistently, than an instructor could on his or her own. The learning analytics system might also be able to identify “non-obvious” factors that correlate with positive or negative learning outcomes, which even an experienced educator might not recognize.

Sound cool?! Problem is, learning analytics tools are costly to develop and productize – and typically they aren’t “one size fits all.”

So who’s doing learning analytics today? Some colleges and universities have begun adding learning analytics tools to their learning management systems (LMS). LMS vendors have also begun adding learning analytics features in their software. For-profit colleges are among the early adopters. Bricks-and-mortar institutions like Louisiana State University and Rio Salado College in Arizona are also beginning to use learning analytics for predictive modeling.

In my next post I’ll present overviews of some of today’s emerging learning analytics applications, as well as ideas for surmounting challenges and more visions for the future.

Featured image courtesy of IvanWalsh.com.