Spaced repetition: designing for the way we remember

 

Our brains are pretty incredible things. We use them to create memories and send instructions to different parts of our body.

A heap of scientific research and study has gone into figuring out how our brains work. Some good: neuroscience in the last 120 years has come a long way, and some bad: ancient Egypt removed the brain before mummification, believing that the heart was where intelligence existed. Countless studies have gone into developing a better understanding of the human mind, particularly over the last 50 years or so.

It’s a nice idea that our brains are supercomputers sitting in our skulls, ready for instructions to carry out as soon as we demand. A nice idea, but not entirely accurate. Our brains work in mysterious ways, so it is important to accommodate how we are hardwired to think, remember and forget.

Unlike computers, every time we are exposed to a sight, smell, sound or thought, we don’t store it permanently until we decide we need to use it. Nor are our brains like pantry cupboards we can place things in and retrieve them whenever we feel hungry.

Running with this metaphor, if our brains were a pantry cupboard, a whole lot of the food that went in would disappear shortly afterwards. And that’s not because you’re raiding the pantry at night to snack!

If our brains were truly a computer, it would be deleting files it deemed unnecessary at will, never to be seen (or thought) again.

While our brains can remember many details throughout a day, it forgets a whole lot more. The song that was playing as we travelled to work. What colour shoes the person who you accidentally bumped into was wearing. What the person in front of you ordered for lunch.

We experience up to 18 hours of moments, thoughts and ideas every day.

That is a lot of information to process and the majority of it is not retained in our brains.

Our brains aren’t computers (or pantry cupboards). We don’t remember everything we experience in a short-term period. Using a method of that description would be ineffective when trying to retain information. Imagine trying to retain as much information as you can the night before a test or exam…

Cramming is a popular method of trying to retain information. All of us (and if you haven’t, you’re lying) have tried to cram as much information into our heads right before we need it. Going over cheat sheets moments before exams or spending long nights before a test with a number of energy drinks, headphones blasting.

But popular isn’t always the best. Its popularity could even stem from a misconception about short-term memory - which will be touched on shortly. Cramming goes against the way our brains are hardwired to think. Cramming is setting yourself to fail; the most ineffective of ways to retain information. It is all well and good identifying what not to do, but why? What’s the alternative?

 

The first step is to understand the way our brains do work.

 

 

There are three stages of memory that our brains go through when exposed to new information. This is explained in this great video by Crash Course. When we are first exposed to a sight, thought or sound, that memory is encoded and placed into our short-term memory. Generally, our brains can only hold around four or five things in short-term memory at any one time. From there, the memory is either placed and stored into our long-term memory, or it simply disappears.

We just simply don’t remember everything that is exposed to us and it is why cramming does not work.

Memories that are stored go into our long-term memory. However, it isn’t just a case of simply picking out a memory on command; it is retrieved when you are in a similar context. You know the times you find yourself in the kitchen thinking… “Why on earth am I here and what am I looking for?”. Only once you retrace your steps back to the bedroom where you originally had that thought will you recall that you were looking for your house keys.

Memories are able to be retrieved, but it isn’t always as simple as recalling on command.

In 1885, Hermann Ebbinghaus, considered the godfather of memory, found that almost 80% of new information we learn is forgotten within just 31 days. This finding led to the term coined as the forgetting curve. The forgetting curve shows how information is forgotten over a period of time when there is no attempt to retrieve it. Cramming information at one time is not effective as it is not helping transfer information into your long-term memory.

 

Some of the research techniques Ebbinghaus used are questionable, but the overall concept of forgetting over time and a need to actively recall is widely recognised.

Some of the research techniques Ebbinghaus used are questionable, but the overall concept of forgetting over time and a need to actively recall is widely recognised.

 

The idea behind cramming is to remember as much as you can right before you need to retrieve it. However, it’s a misconception that cramming uses short-term memory. Information in our short-term memory “decays completely and is lost within a period of about 30 seconds”, so cramming then becomes an inefficient technique to store information in our long-term memory.

But if cramming goes against the way our brains are hardwired to think, how do we place memories into storage and ensure they don’t just rot away?

 

Spaced repetition

To leverage the way our brains naturally store information, the way we expose ourselves to information must be completely different to cramming. We should be exposed to information repeatedly, but instead of all at once, the exposure occurs with large breaks in between.

Ebbinghaus found that repetitions that are spread out over time allow those memories to form strongly instead of being massed together.

 

spaced repetition pure learning

 

A study in 2006 confirmed that spacing out learning events by “a period of at least 1 day, rather than concentrating all learning into one session, is extremely useful for maximizing long-term retention.” The study demonstrated a benefit from the learning being distributed across weeks and months as compared with a few hours or just a single day.

So how long should these spaced out periods be? There is no one-size-fits-all period, but the evidence suggests that this time period should be more than one day if the information needs to be retained for a number of months or years.

These periods between each event of practice need to increase exponentially. That is to say the first spaced out period should be followed with a larger period to help reinforce the knowledge, and so on. As the gap increases, and as long as the retrieval is challenging (yet manageable) to the individual, there is an increase “the probability of successful trace retrieval.” By having challenging practice at longer intervals, the memory connection is reinforced and made even stronger.

What does this look like in a practical sense for learning in the workplace?

It means that one-off learning ‘events’ are not going to entrench knowledge sufficiently. Instead of sitting at a computer for an hour trying to remember everything that you see, the learning needs to be spread out over weeks and months. There needs to be multiple instances of learning to embed the knowledge.

Learning strategies need to be meticulously planned to ensure any learning is embedded, and that learning is not considered as a one-off event. There needs to be follow up practice that is challenging to the individual.

This can be in the form of a campaign-based approach, a spaced out blended learning program or a much larger and planned out ecosystem of learning and performance.

No matter how you’re tackling workplace learning, you can’t rely on putting people through one training session and expecting long-term meaningful change.


Some light reading that expands on the science and studies behind this concept.

Atkinson, R. C., & Shiffrin, R. M. (1968). Chapter: Human memory: A proposed system and its control processes.

http://cogs.indiana.edu/FestschriftForRichShiffrin/pubs/1968%20Human%20Memory.%20Atkinson,%20Shiffrin.pdf

 

Ebbinghaus, Hermann (1885). Translation of Memory: A Contribution to Experimental Psychology
http://psychclassics.yorku.ca/Ebbinghaus/

 

Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis.

http://www.evullab.org/pdf/CepedaPashlerVulWixtedRohrer-PB-2006.pdf

 

Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2009). Distributed Practice Over Multiday Intervals

http://wixtedlab.ucsd.edu/publications/wixted2/Optimizing_distributed_practice.pdf


 

 
Beau Bardenhagen