Outrageous Conclusions in Big Data and Learning

By A.D. Detrick

Learning professionals have been trying for decades to integrate personalized learning into instructional design. The concept of “learning personalization” is a powerful way to reach learners with the best learning tools available to them, based on their learning style. The challenge in the past was that despite our good intentions, we simply lacked the tools to implement those personalized elements appropriately.

Enter big data and its incredible value for quantifying exactly how individuals learn best.

Today, the data we collect provides insight into where, when, how, and why people engage with training. And we can collect data on an individual level, not just at the audience level. All of that data helps in identifying optimal paths to learning. The concept of learning personalization is simply one of the most important benefits of using big data for learning.

Fixing the Fatal Flaw

In the earliest days of instructional design, learning organizations tried to identify the best ways to present information to learners. Once we had identified the most common learning style among the entire audience, we could design for the average. We had an idea that people had an ideal or innate learning style – but that research turned out to be deeply flawed. Those findings were based on a survey methodology that allowed people to self-identify the ways they believed that learners learned best. Needless to say, it’s been thoroughly debunked in recent years.

Today, we know much more about the way the human brain is wired. We know that we’re hard-wired for exploration: discoveries are literally rewarded with chemical flashes inside our brain’s reward centers.

That these flashes can often occur in similar learning scenarios indicates a preference for learning, not a defined style. After all, preferences are not absolute. I may prefer coffee in the morning, but in the hottest days of summer, my preference shifts to iced tea. Learners experience the same shifts in preference as they search for information to do their job. We may like certain learning modalities because they’re familiar to us, or easy to follow. But regardless of our perceived preferences, all of us are hard-wired to search for information – to look for that zing of a reward upon the discovery of new information.

Historically, our training efforts have not been designed to build upon that system. We’ve stored the information on how to do a job in standardized formats – manuals of policies and procedures for every situation. And we’ve trained people with the most average style, in effect limiting the scope of exploration and dramatically reducing the possibility of a discovery reward. Training became a process of memorization and repetition, which we forced on learners who were hard-wired for discovery and reward. Whoops.

The Big Data Zing

Back to big data. Tools now allow integration of multiple channels of data. The policies and procedures manuals can remain. But, today, we can use data measurement tools to easily track how frequently they are accessed – and even track access by page or paragraph. We can create and store user-generated content (short instructional videos, blog posts, job aids) and track its access. We can even track internal communications over channels like IM (instant messaging) and Chat. And we can mine that data to find each individual’s preferences. We’re no longer tied to designing for the averages. Now, we can design personalized experiences for each employee. We can identify those who prefer video learning and those who prefer to message a peer for resolution. We can even identify circumstances that might cause a learner’s preferences to change, and adapt our recommendations to give them the modalities that might best create a discovery and reward situation.

By leaning on big data for learning personalization, we not only expand the volume of learning opportunities, but we can deepen the effect of that learning. We no longer need to be constrained by a lack of information which forces learning on our audience in a manner they may not prefer. Multiple modalities of the same information no longer means the training department is unfocused or inconsistent. By leveraging data on learner preference, we can create a constantly changing learning landscape, optimized for each individual, to allow them to explore, discover, and be rewarded for their discoveries. It’s a new way of imagining how big data can help design the most robust learning solutions.

Subscribe to this blog and receive email updates when we publish a new article.