I was never explicitly taught about memory in my eight years of teaching. This is no one’s fault as we are all doing the best we can. However, having starting reading books and blogs there is a whole world of education research still untouched by many. One important area that I feel all teachers should know about is that of memory and how memory works.
We all want our students to remember stuff and I am sure we can all empathise with each others frustration at the students knowing stuff during lessons but completely forgetting it when it matters, during tests. Understanding how our memory works is key to tackling this all too common classroom occurrence.
In my last post we briefly explored Ebbinghaus Forgetting Curve and we discussed our ability to retain information through spaced retrieval. To understand this in greater depth we must explore our working memory and long term memory.
The first area to consider is that of our Working Memory (WM). Research into our WM has shown that it is a finite resource, with some researchers claiming we can only hold up to 7 (plus 2) ‘chunk’ of information at any one time, some recent research has suggested that the number could be as low as 4 ‘chunks’. Our WM is where we process information from our current environment and also draw upon knowledge from our long-term memory. As mentioned our WM is a finite resource but it is always active and processing information. To reiterate, our working memory is always full, it is taking in everything in our surrounding. As David Didau writes:
Working memory is synonymous with awareness. It is the site of conscious thought. The act of paying attention, of reading these words, of listening to your children complain about how much homework they’ve got to finish for Monday morning, fills up our working memory. In practical terms our, our working memories are always active, even when we are focussing on something in particular. We’re constantly absorbing and processing sensory data from the world around us.David Didau, Making Kids Cleverer.
This is where Sweller’s ‘Cognitive Load Theory’ comes into play for us teachers. Often we ‘overload’ our students with too much information which produces too much cognitive load. To lighten this cognitive load our students have to have acquired knowledge in the long-term memory, referred to in literature as schemas (schemas are basically folders of knowledge on one topic, the more the folder is filled with knowledge the lighter the load on working memory for that particular area of knowledge).
As you can see in the graphic our working memory fills up and we can either learn the material by storing it in our long-term memory or forget it. If our WM is filled with too much cognitive load then whatever else is added will most certainly be forgotten so it is worth learning more on Cognitive Load Theory.
Despite the apparent bottleneck of our working memory there are strategies that we can use to overcome cognitive load. Firstly, having a vast store of knowledge in our long-term memory in the form of schemas will certainly help, in simple terms – the stronger the schema the lighter the load on our working memory (this is basically why experts make some things look so effortless and novices struggle so much).
Baddeley and Hitch’s Working Memory Model (WMM) is one of the most robust theories in cognitive science and gives us further insight into working memory. The Central Executive (CE) acts a bit like a supervisor or night club bouncer. As information tries to enter the ‘club’ the CE focuses attention on the information and decides which one to attend to, or to continue the analogy which one gets into the club and which information gets the good old ‘not tonight pal!’. It’s important to note that the CE is entirely under our control and is more of a subconscious function. Every teacher can speak to losing a classes focus when a wasp enters the room!
The Phonological Loop (PL) deals mainly with speech and other types of audio. This is where we store verbal information, up to about 2 seconds before it is overwritten and new information comes in. We either move it to our long-term memory or forget it.
The Visuo-Spatial Sketchpad (VSS) briefly holds visual information and the spatial relationship between things. Imagine the light goes out in your room, this is where you store the memory of where all of your clutter is so that you don’t trip up.
The Episodic Buffer (EB) was added to the WMM in 2000. It was added because there was evidence to suggest we needed a component to combine the information in the WMM stores to what we already know – our prior knowledge stored in our long-term memory. This shows the importance of a vast array of knowledge in our long-term memory. The more you know and the larger your schema is for a topic, the lighter the load on your working memory.
The role of LTM in helping working memory is well established and very easy to demonstrate (e.g. – compare the retention of a random sequence of letters – DPL OAM IGGB – to a sequence containing meaningful ‘chunks’: DOG PIG LAMB)Taken from ‘What every teacher needs to know about psychology’, Didau & Rose
One of the key things to note about working memory is just how limited it is.If you are distracted while trying to process something you will lose the information you are trying to process (think about what this means when the damn wasp flies in.) We also can only deal with a small amount of information at any one time as suggested by Miller’s ‘7 plus 2’ chunks from his research in 1950.
This is why it is so important that teachers know about Cognitive Load Theory and Dual Coding Theory to help them combat the limited working memory of their students. Dylan Wiliam said that cognitive load theory is the “single most important thing for teachers to know” However, to really help our students we should be working hard to get the information we teach into their long-term memory.
First, we must note that our memories are invisible to us and there isn’t any consensus as to where exactly our memories are stored but we know enough that our long-term memory is vast and perhaps even limitless and the more stuff we have in there the easier it is to learn as the working memory load will be reduced. Learning has been defined as “a change in long term memory” by Kirschner, Sweller & Clark. If we run with this then, it is our acquisition of schema that fills our long term memory.
Storing memories is about making links and connections between our experiences in a vast network of related concepts and contexts. These links and connections are referred to as ‘schema’. As mentioned earlier a schema can be though of like a folder in your laptop that gets filled with the relevant knowledge in one given topic. An example of a schema in action is as follows:
A frequently used example is going to a restaurant. The schema for getting a table, ordering food and drink, and paying for the meal makes visiting a new restaurant for the first time, even in another country, a pretty straightforwards process, as we deal with new situations by linking them to things we’ve encountered in the past.David Didau, Making Kids Cleverer
Our long-term memory isn’t a single storage unit and psychologists tend to divide it into to separate but interlinked systems: declarative memory and non-declarative memory.
Non- declarative memory is a catch-all term for everything that may exist in our long-term memory that we are unable to put into words. An example of this is your ability to read this sentence and understand the phoneme-grapheme correspondences required to read this, you just know how to do it (even though it was once a challenging and hard learning experience). Other procedural skills like tying your shoelaces, walking, swimming or cycling are features of non-declarative memory.
Declarative memories are the memories we can declare: “Cristiano Ronaldo plays for Real Madrid”, “they are 30 years old”, “pythagoras theorem is a2 + b2 = c2” and so on. Declarative memory can be either episodic or semantic.
Episodic memories are those of experiences and specific events, how you felt at during those events. We can often replay events in great detail using our episodic memories. Whereas, Semantic memories are a more structured record of facts, concepts and meanings. Episodic memories are mainly context dependant but semantic memories are more flexible and can be applied across a range of contexts.
The two systems, episodic and semantic are linked in several ways. Semantic memories can become ‘stand alone’ memories but they are often derived from a specific episodic memory. In terms of teaching an episodic memory could be that of a particular lesson and the semantic memories are the facts, key terms and concepts of that lesson. Quite often our students can recall episodic information from a lesson but struggle with recalling the semantic information.
Understanding episodic and semantic memory can help us, as teachers, understand why our children oftentimes can’t recall what we teach them. They remember the episodic memories of lessons – messing about with friends, Mr Murphy’s horrible breath and being given detention for incomplete homework. In order to make our semantic memories stronger we must retrieve factual information often which will allow us to retain our learning over the long term. Which is why retrieval practice really is an important pedagogy to undertake.
If we don’t retrieve the semantic memories, when asked “do you remember when we learned about plate tectonics?”. The students might reply with “oh yes i do remember” but they may be recalling the episodic memory and not the semantic memory and unless the teacher digs deeper with further probing questions the student will have the illusion of knowledge and perhaps be relying on the familiarity effect, with no change in their long term memory.
There have been great studies that have revealed the links between semantic and episodic memories. The most famous of these is by Elizabeth Loftus and John Palmer. They showed participants of their study a series of films involving car collisions and found that estimations of the speed the car was travelling could be manipulated by changing the verb used in their question. Where participants were asked “about how fast were the cars going when they hit each other?” they gave lower speed estimates when compared to participants who were asked “about how fast were the cars going when they smashed each other?”. The change in language appeared to create a ‘fact’ about the collision which influenced the memory of the collisions they witnessed.
As mentioned earlier a ‘schema’ is like a big folder with interrelated concepts and contexts and is assembled of non-declarative and declarative memories. Some of what we remember is semantic, some is episodic, but they are all stored somewhere within our brain.
Didau, David and Nick Rose (2016) What every teacher need to know about psychology
Didau, David (2015) What is everything you knew about education was wrong?
Didau, David (2019) Making Kids Cleverer