4 Steps to Bring Learning Analytics to Online Education

Have you ever wondered whether your students are really learning what you’re teaching in an online format? Or if they’re struggling and why? Testing and grades can tell educators something about all that, but only “after the fact.”

What if you could get actionable feedback during the delivery of your course that could allow you to adjust the curriculum and provide supportive feedback on an individual basis to help learners be more successful?

While an online classroom might not offer the same face-to-face opportunities for interaction as a physical classroom, it actually has some advantages when it comes to assessing and impacting student performance.

That’s the premise behind learning analytics, at least as it’s envisioned today in terms of its applicability to online education. In this emerging area of applied research, technologists and thought leaders are envisioning ways to reveal how engaged students are, how much they’re learning, and how to enhance the lessons going forward.

How does all that work? That’s precisely the question that this terrific infographic from the folks at Open Colleges was created to answer (here’s a link).

Learning Analytics: Leveraging Education Data – An infographic by the team at Open Colleges

To capture learning analytics in an online classroom, four basic steps need to happen:

1)   You need to collect large volumes of data from the learning environment. Other channels could also be involved, like social media sites. Some of the data you might want to capture about individual learners would include: amount of time spent in the environment; time between logins or time since last login; and grades and other data in other areas of study.

2)   Next, you need to analyze that data to derive meaningful conclusions. For example, a “profile” of a struggling student (as derived from large volumes of data on lots of other students) might include factors like: low time spent on the site versus colleagues; a long time between logins; below average performance learning about similar topics; and answer patterns on quizzes indicating a fundamental lack of understanding (Socrato can do something like this for SAT/ACT test prep).

3)   Once the learning analytics “system” or “engine” identifies a problematic constellation of behaviors, it could “personalize” the learning environment to provide extra help. For example, if a student appears to be giving up too soon on a math problem, the learning environment could display clues, prompts, feedback and encouragement — in real-time.

4)   The more data you get across more and more interactions, the more educators can track what works and fine-tune the learning analytics system.

It might not be too long – maybe just 3 to 5 years — before individual students can learn online with custom curriculums that fit their learning style, strengths and weaknesses, and more. Why haven’t learning analytics emerged already? Cost, technology and privacy issues are all still being worked out. Learning analytics “done wrong” could lead to misinterpretations or worse.
What does this emerging learning analytics paradigm mean for teachers and tutors? Besides instruction, teachers will also facilitate interaction with the learning analytics system, and analyze the results.

“Grading” per se might become less of a factor in teachers’ lives, as learning analytics systems could enable “peer grading” and even “self-grading” based on data and correlations with current and past performance.

How do you envision learning analytics impacting your educational environment, tools or techniques?

Featured image courtesy of dougclow.