Actively using memory to enhance learning

When you consider a team sport, like soccer, individual players like Messi are sure to come to mind. But, when it comes down to it, winning is a team sport. Even though every team might have one or two all-star players, those individuals are not winning the games by themselves. Instead, when they play, they influence activity on the entire field. Every action one player takes affects where the other team members run, who they block, or when they pass the ball.

Memory is not an island

Memory is a team sport too. Our brains aren't filled with a bunch of isolated facts. One thing that makes us so intelligent is our ability to make connections among seemingly unrelated pieces of information. This is the heart of memory. At the same time, we don't form associations between every piece of content we encounter. Our brains are relatively selective at deciding what gets gelled together. How does it do that?

One idea is that we have both "active" and "inactive" memories at any given time. We have accumulated vast amounts of information over our lifetime, but we don’t need to have instantaneous access to all of it all of the time. Just imagine how confusing that would be -- we wouldn't be able to function! So, our brains can efficiently store this information and allow us to gain access to it when we receive some kind of trigger that queues up the memory. When this happens, we can think of that memory as being "activated" so we can use it to make decisions and guide our behavior.

Testing accelerates learning

'Once you stop learning you start dying.' -- Albert Einstein

When you were in school, you probably remember a never-ending cycle of study-test-repeat. Once you finally graduated, you didn't have tests constantly looming over you anymore, but that doesn't mean you stopped learning new facts and skills. Whether its learning how to use a new app or speak a foreign language, effective learning strategies are highly valuable. There are so many great tools out there to teach people new skills. Web-based programs like Duolingo and DataCamp are popular tools that drill language and computing concepts. What makes these tools so effective?

These web-based programs emphasize one thing: active engagement. Rather than simply reading material and repeating it over and over until its drilled into your head, effective learning programs get you to actively engage in the material. This active engagement can be memory retrieval, which involves actively remembering some specific content you studied. Or, it can involve asking you to creatively apply concepts to new situations. Both of these approaches require you to employ some agency during the learning process. Agency does not apply to situations where you simply go through the motions without any real thought. It requires voluntarily remembering specific content, making a choice, or combining information in a way that creates something new.

How does actively remembering influence behavior?

I have investigated how actively remembering a specific event influences how we engage with other content in the environment. This research has focused on how we dynamically use memory to guide our behavior and influence how we learn new information. The idea is just like the soccer player analogy from above. How does moving a soccer player off the bench and onto the field influence the entire team dynamics?

Remembering promotes learning

I tested the idea that remembering a specific event would help people learn other information more efficiently. I wanted to see if actively engaging in remembering puts you in an ideal state to gather new information. The trick was making the new information totally irrelevant to the remembered event, and getting subjects to encounter the other content immediately after remembering. This would allow me to test the idea that actively remembering creates a optimal brain state to learn new information, even if to-be-learned content has nothing to do with the specific event that was remembered.

For this task, subjects studied object-location associations. After a delay, subjects engaged in an Active or Passive condition. In the Active condition, I asked subjects to remember each object's location and move it to that spot on the screen. In the Passive condition, I asked subjects to simply move the object to a yellow box on the screen. In the Active condition, subjects had to voluntarily bring the object's location to mind. On the other hand, in the Passive condition, subjects did not have any agency to decide where to place the object. Immediately after the subject moved the object, they encountered a picture of a face they had never seen before.

Later, I tested subjects' memory of the faces. The hypothesis was that subjects would remember faces better if they encountered them right after they actively remembered an object's location rather than right after they passively moved an object to a location.

Interestingly we found just that. Memory was better when they encountered the faces immediately after actively remembering, rather than passively moving, an unrelated object (Bridge, 2014).

I hypothesized that actively remembering a specific event produces an active brain state, that makes learning easier. To test this idea, we also recorded brain activity while subjects studied the faces using scalp EEG. With this data, I could see if brain activity differed when subjects studied faces in the Active vs. the Passive condition.

The EEG data looked remarkably different in the two conditions. When subjects remembered faces in the Passive condition, we observed a brain state that is very similar to other studies investigating neural correlates of successful learning. Neural activity that predicted later memory was apparent more than half a second (~600 ms) after subjects encountered a face -- an effect most researchers infer as reflecting a memory formation process. These results suggest that subjects are usually in a passive state when they learn information in typical computer-based experiments.

On the other hand, we found completely different results in the Active condition. Neural activity that occurred immediately after subjects encountered the face (~100 ms after face onset) predicted later memory for faces in the Active condition. These unique brain correlates of successful learning in the Acive condition suggest that something special was happening when subjects learned new information in an Active state. These data confirmed the idea that actively remembering a specific event changes the current brain state, making it an ideal time to learn new content.

Conclusion

Actively using memory is extremely powerful tool to learn new information. It enhances attention, and promotes rapid integration of new content into our existing frameworks.

Main Points

  • Actively remembering produces a unique brain state, making it ideal to learn to content quickly and efficiently.
  • Actively remembering specific memories influences behavior and determines what we pay attention to.

GitHub