Back to the blog

The fickle nature of our working memory

By Tom O'Donahoo on 6 May 2020NSWstaffroomUKstaffroomInnovation

In the early 2000s, a group of researchers at Japan’s Primate Research Institute (PRI) began running a series of working memory tests on chimpanzees in order to establish how their memory might differ from humans. As some of our closest relatives in the animal kingdom, knowing what separates us from primates helps us understand more about ourselves and our evolutionary history.

However, something relatively unexpected happened.

As it turns out - a chimp’s short term memory is scarily good. When there’s a tasty peanut on the line, the chimps consistently beat their human counterparts at remembering basic visual sequences. Take a look:

What educational psychologists hope to understand about memory

Human memory has a long history of interest in academia, and much of that research has focused on how information is transferred from our short term or ‘working’ memory into our long term memory.

A memory model showing how information and sensory input is processed into working and long-term memory.
A working memory model

When we take in sensory information like sight or touch, it enters into our sensory register. From there it moves into our working memory before it’s encoded and transferred into our long-term memory where we later retrieve it from. By better understanding this transfer process, educational psychologists hope to understand how we can efficiently retain new and critical information.

World-renowned educational neuroscience consultant David A. Sousa examines how brain research can improve teaching and learning. In his book, How the Brain Learns, he breaks down two questions the working memory uses to determine what information to transfer to long term memory. First, ‘does this make sense?’, and second, ‘does this have meaning?’ To increase the probability that students store information, we need to make sure we’re delivering it with both sense and meaning. Taking this research as inspiration, education should be tailored to individual students’ understanding every step of the way, making sure that the information we’re asking them to commit to memory is always relevant.

Efficient strategies for accessing long term memory

Although there are endless strategies out there, there's a substantial body of research suggesting that we can increase the efficiency of retaining information in our long term memory if we:

  • Make the information as simple, clear and digestible as possible. This reduces the cognitive load on our very limited working memory.
  • Revise the information at carefully spaced, approximately exponentially growing intervals

Given this, it’s been shown that teachers can help improve the efficiency of encoding long term memory through reducing the load on their students’ working memory through carefully constructed explicit teaching and scheduling regular time for revision.

Tailoring the approach for individual success

Let’s preface this section by admitting that keeping on top of students’ retention is no mean feat. Let alone calculating how far apart revision intervals should be, and how this should all be tailored to each individual student.

At Atomi, we use Artificial Intelligence to help teachers predict how their students have retained information over time. This technology uses data from millions of test scores, taking into account factors such as the student’s individual history, the difficulty of the topic, and how and when they’ve interacted with it before.

These predictions help teachers identify who might need help, when students need to revise, and how students are tracking leading into exams and assessments. We use these predictions to create personalised revision recommendations for each student, targeted at what’s most important to their individual success.

Overall this can help teachers make the complicated process of remembering relevant content significantly more efficient, transparent and easier to manage.

Hopefully, with improvements in technology, we’ll one day be less reliant on our fallible human memory. Maybe all facts, figures and what we wanted to pick up at the grocery store will be just as easy to recall as a sequence of digits for a chimp. But until then we’ll have to keep working hard to better understand the fickle beast that is our human brain, and how to best utilize what we’ve got.


Liked what you read? Check out more awesome articles:

Atomi Brainwaves: Jocelyn Brewer on Digital Nutrition

Sharing secrets: The Atomi guide for creating engaging video lessons

How to spot students falling behind in the remote classroom


Try Atomi for free and receive regular updates from our blog.

Learn More