First look: Brain Data and Online Learning

We’ve all had the experience; sitting in a classroom, the sound of the humming HVAC system in the background. The smell of old wood, musty carpet, and industrial cleaning supplies combined. You’re trying to focus and follow along as best as you can, when all of a sudden there is a new image on the overhead projector, or your teacher has just said a complex, unfamiliar word. A wave of anxiety comes over you. You feverishly take notes and flip through the pages of your pre-reading.

You look around the classroom to see if anyone else looks confused. All your classmates are either mastering this material or they’re mastering the art of disguise. Straight faces and only the sound of someone clearing their throat as your teacher trudges forward. You think you’re all alone. Everyone else gets this but you! And then the teacher continues, labels appear next to the new image or worse yet, you hear “but don’t worry about that now, we’re going to cover that in a separate lesson”.

You move forward, you realize you’re learning about something you did prepare for. And in the latter case, you renew focus knowing the unfamiliar information was irrelevant for today, at least. You look around, everyone else seems relieved; you weren’t alone. Now, imagine this entire scenario playing out as you sit at home alone. Sitting at your makeshift desk. Confined to online learning because of impacts we’re all experiencing due to the global pandemic.

UNESCO is tracking the impact of the COVID-19 global pandemic on students world-wide. At the height of the impact in May 2020, more than 1.4 billion students were affected by school closures. That’s 84% of all primary, secondary, and post-secondary learners. And that doesn’t even consider the volume of professional education and corporate training that moved online in the past 8 months. Even before COVID-19, there was high growth and adoption in education technology, with global edtech investments reaching almost US$19 billion in 2019 and the overall market for online education projected to reach $350 Billion by 2025.

Today choice abounds in online learning; from formal academic institutions to private training providers, to individual contributors producing content on YouTube. Options for education are changing rapidly, and prospective students don’t currently have a mechanism to understand, from the perspective of neuroscience, which learning platform or content is going to do the best job in teaching them.

Viruses are top-of-mind around the world right now. The topic of viruses is covered in high schools, tertiary academic institutions as part of advanced science degrees, and online learning options such as private training providers and individual learning content developers. So, this is where our journey to understand the impact of online learning material on our brains begins.

We wanted to understand similarities and differences across different education brands on effectiveness measures such as neural engagement, attention, connection, encoding to memory, as well as other factors like cognitive strain.

We identified three segments of online learning content from a variety of producers; a top search result from YouTube when inquiring “what is a virus”, learning material from a credit course at one of the oldest Ivy League schools in the United States, and learning content produced by a private, non-credit educator that has experienced explosive growth over the past decade.

The content was tested through Brainsights, a Toronto-based neuromarketing network, that connects neuro technology with human participants to record the unconscious brain activity and directly examine how the brain is responding to content. Brainsights measurement platform is powered by Electroencephalography (EEG) brainwave readers, and focuses on three core metrics; attention, retention, and encoding.

Attention, connection, and encoding are measured from relevant frequency bands within EEG data. The Attention metric is reporting on overall alertness and attention being paid to the content, the Connection metric is reporting how deeply the content resonates emotionally, and the Encoding metric is reporting the extent to which the content is being committed to memory. Overall Neural Engagement is a composite score derived from Attention, Connection, and Encoding combined and intended to provide a summative vantage point into how the content is engaging an audience overall and moment to moment.

Through combined expertise in instructional development, neurotechnology, and data analysis, the raw EEG data was analyzed to identify insights that can aid in future learning design and development.

Relative Neural Engagement

Neural Engagement data provides content creators with the ability to understand how impactful their content is moment-by-moment and overall. When analyzing the overall neural engagement scores of each of the pieces of online learning material, we found that there are relatively small differences. The learning content sourced from YouTube was overall 3% above benchmark, whereas the learning content of the private educator was 4% below the benchmark. The university course learning material was aligned to the benchmark score. With only a 7% spread overall, the ways in which the different online learning content performed moment to moment paints an insightful picture for students and teachers.

Reinforcement Moments

As we investigate moment-to-moment engagement within each of the pieces of learning content, it becomes clear that there are instances where the process of connecting with students and encoding information to memory is more successful and efficient than others.

What we see in the brain data is an increase in encoding to memory once all of the relevant information in both audio and video format has been presented and the instructor takes a moment through audio to repeat the concepts after the corresponding image has been completed. The teaching approach of delivering audible information as the instructor uses a digital canvas to draw visual elements is common, but can produce lags between the production of audio commentary and the production of corresponding visual elements.

In the timeframe for encoding data depicted here, the teacher has just completed delivering a series of successive audio and visual cues and has left the visual component static as they take time through audio to reinforce the concepts presented leading up to this point. Throughout the data of all content providers, we saw several relative increases in encoding to memory that were associated with moments when the teacher was reinforcing or reviewing information presented earlier in the lesson.

Storytelling and Efficiency

Studying neuroscience measures for online learning material where the three approaches were covering the same concepts indicates which online learning content design is most effective. The YouTube implementation delivered content in a clear progression from simple to complex. Audio and visual cues were tightly aligned through this period of time, along with an image that built up gradually as the material was being presented to students. This allowed students to focus on key elements of the lesson in the moment. As the story was being told, we see the highest levels of connection and encoding from the YouTube video. It's important to note that the teacher was not simply reinforcing information that had been previously presented. In a more efficient approach to online content design and delivery, high levels of connection and encoding are achieved as the material is presented for the first time.

Instructional developers and teachers should consider important elements in the storytelling of their online learning lesson plans. Mapping information delivery from simple to complex, with well timed audio and visual elements, and carefully selected labelling in visual elements to draw focus to key terms, has shown here to contribute to improved connection to content and encoding to memory.

Pre-training and Continuity

Looking to another period of time in the learning material, the data highlights the importance of continuing to build on the story you have been telling. In this section, the same YouTube video introduced a new animation. The animation contained new visual representations of cells and viruses without labels or audio description that might have helped the students understand what they were seeing. Encoding levels seen earlier dropped off significantly, signaling that students were no longer engaged in the content or how it connected with concepts presented earlier.

On the other hand, the university and private educator content presented material with familiar images and conversational tone and experienced relatively sustained levels of encoding throughout this topic.

Application

The findings of this study are aligned with elements of the body of research in cognitive learning theory. Design principles around presenting corresponding words simultaneously versus successively, labelling visual elements to highlight key concepts, and introducing material from simple to complex have shown to have a positive impact on attention, connection, and encoding in the brains of students. As with all content developers, teachers will benefit from learning and applying concepts that avoid the introduction of extraneous information toward creating more efficient learning content. As we move forward through rapid adoption of online learning, an opportunity exists to leverage machine learning in the identification and assessment of online learning content that is likely to be the most effective based on neuroscience. Along with the advancement of Electroencephalography (EEG) brainwave readers for in-home use, we envision a future where the power of neuroscience is in the hands of individual students. Learning is, after all, a highly personal and integrated experience in all of our lives.