Cognitive load theory and the design of ITT curricula

Douglas Fairfield and Jenny Griffiths, Teach First, UK

I have a very clear memory of a physics lesson that I taught to a GCSE class in the first few weeks of my initial teacher training. It did not go according to plan. I had studied biomedical sciences at university, and the last time I had studied physics was in my GCSE year at school. But there I was, trying to explain why you could drop a cannon ball and a tennis ball from the same height at the same time and they would hit the ground at the same time, despite the cannonball being significantly heavier. I didn’t really know the answer myself to be honest. I had learned the reason the night before and it had made sense at the time, but now at the all-important moment my memory was failing me. Part-way through my fumbling explanation, a student spoke up. ‘But Sir, if falling objects accelerate at the same rate then why does a feather fall so slowly?’ It was a very good question and I didn’t have an answer. Unfortunately, it was around then that I noticed that several of the students at the back of the room were talking quietly among themselves, and I had no idea how long they had been off-task. Why hadn’t I noticed that they were talking? Why couldn’t I remember how to explain simple physics? What should I do about the chatty students at the back of the room? How should I respond to the falling feather question? What should I deal with first? It was infuriating and overwhelming. My impulsive behaviour intervention strategy – ‘Why are you talking when I’m talking?’ – did not have the effect that I had hoped for (to be honest, I’m not even sure that I had thought through what effect I was expecting).

The reason I remember this lesson so clearly is because it made me realise that there was too much going on in the classroom to be able to think about at once. That evening I decided that I would need to invest time in practising. I practised: giving clear and concise explanations, the questions I planned to ask, embedding student thinking time after asking questions and calling on pupils for their answers after appropriate thinking time. This practice involved scripting out the precise words and phrases that I intended to use in my explanations and key questions, before practising these scripts outside of the classroom environment as if I was in the live lesson. Then I read up on Doug Lemov’s ‘least invasive interventions’ and began to practise ‘non-verbal’ interventions, ‘anonymous group corrections’ and ‘private individual corrections’ (Lemov, 2015). I would imagine situations in which I would need to use each type of intervention and practise how I would respond. This probably wasn’t the best order to do it in, but I didn’t have anyone offering me a better sequence of learning.

Seven years later, I am confident that I could manage that lesson with ease, so why did I find it so challenging then? This can be explained by two effects of cognitive load theory (CLT): ‘element interactivity effect’ and the ‘narrow limits of change principle’ (Sweller, 2019). These effects suggest that the more the elements of new information that need to be processed by the working memory and the greater the number of interactions between these elements, the greater the cognitive load on the working memory. This is known as ‘element interactivity’ (Lovell, 2020). The narrow limits of change principle refers to the ‘severe limitations of working memory when processing novel information’ (Sweller, 2019, p. 273). While working memory is not fixed, it is finite when processing new information. I struggled with my physics lesson because there were too many elements of new information at play, all adding to my limited working memory capacity and resulting in cognitive overload.

So why can I manage similar lessons with relative ease now? One of the key terms in the narrow limits of change principle is ‘novel information’. While the working memory is finite in its capacity to deal with new information, ‘there are no known limits when familiar, organised information from long-term memory is processed’ (Sweller, 2019, p. 273). The ‘environmental organising and linking principle’ (Sweller et al., 1998) of CLT claims that the limits of the working memory disappear when information is retrieved from the long-term memory. As an initial teacher trainee (ITT), I lacked the rich depth of domain-specific knowledge relating to teaching and learning embedded within my long-term memory. It is these deep reservoirs of domain-specific knowledge that Sweller et al. describe as ‘the major, possibly sole difference between novices and experts’ (2011, p. 21).

It is important to note that there is variation between individuals in their ability to process information in their working memory (Meinz and Hambrick, 2010). However, because of the environmental organising and linking principle, this variation in working memory capacity is overshadowed by the large stores of domain-specific knowledge stored in the long-term memory. The implication for teaching effectiveness is that, for any individual teacher, a large store of relevant knowledge in the long-term memory is more valuable than a larger working memory capacity. This is the challenge faced by trainee teachers: they need to rapidly develop their ‘store’ of relevant knowledge over time.

Berliner argues that the development of automaticity and routines for repetitive operations is a key characteristic that marks an expert teacher (2004). Key to this development of automaticity is ‘situation-to-action pairs’, a large collection of situations and associated actions stored in their long-term memory. These situation-to-action pairs allow an expert to recognise a situation, select an appropriate strategy and execute that strategy without adding to the intrinsic load of their working memory capacity (Lovell, 2020). We see this all the time in experts in a range of different domains. The classic example is chess masters storing vast numbers of different board configurations along with the most successful moves for each configuration in their long-term memories (Simon and Gilmartin, 1973). A novice chess player does not have these situation-to-action pairs within their long-term memory and so finds it harder to decide what move to make. In a highly complex game like chess, it is likely that the novice makes a sub-optimum move.

How is this relevant to ITT curriculum design? As explained, the element interactivity effect can lead to cognitive overload in new teachers, overwhelming them with the complexity of teaching. However, as the amount of domain-specific knowledge relating to teaching stored in their long-term memory increases, they build more situation-to-action pairs, allowing them to respond automatically to classroom situations, thus developing expertise. This automaticity reduces element interactivity through the environmental organising and linking principle, i.e. concepts and procedures that comprise many interacting elements can be stored in long-term memory as one single element (Sweller et al., 2019). One of the best things that we can do for new teachers is to design a curriculum that prioritises the development of rich domain-specific knowledge in their long-term memories. In other words, an ITT curriculum needs to develop their expertise as rapidly as possible.

Understanding CLT helps us to design our ITT curriculum by considering the ‘isolated elements effect’ (Sweller et al., 2019). Trainees can be supported by breaking down the complex concepts and procedures to learn separately, before seeking to integrate them into the whole as a ‘single element.’ This can be done through deliberate practice of key pedagogical skills, not only in the classroom, but also outside of it. My lessons began to improve when I started scripting and practising my explanations; through practice, they became routine and with that routine I reduced the strain on my working memory. My lessons became better still when I scripted and practised the use of non-verbal interventions. As these also became routine, I reduced the strain on my working memory even further. With time, I was able to integrate my explanations with the use of non-verbal interactions as one automatic element, but I was only able to do this because I practised them in isolation first. Since teaching is a highly complex profession with many interacting elements, embedding deliberate practice of essential ‘gateway skills’ is an important part of the design and delivery of the ITT curriculum. The importance of deliberate practice outside of lesson time cannot be overstated. Trainee teachers only practising new skills in the context of live lessons is like expecting a lead guitarist in a band to perform a new song for the first time in the middle of a live gig. Deliberate practice outside of lesson time offers an important opportunity to avoid cognitive overload and therefore improve performance.

Instructional coaching can make a key contribution to supporting the delivery of an ITT curriculum, as it provides the opportunity for deliberate practice outside of lesson time with the support of an expert colleague. A recent metanalysis (Kraft et al., 2018) has highlighted the potential benefits of incorporating coaching into professional development programmes, with a pooled effect side of 0.49 standard deviations (SD) on teacher instruction and 0.18 SD on student achievement. When I started teaching I often felt isolated, constantly doubting whether I was practising the right things in the right way and the right order. Was I focusing on what the students needed? Trainee and early career teachers need expert colleagues as mentors to guide them, to model and explain teaching actions for them to practise. They need a mentor who observes and gives constructive feedback in a non-evaluative manner, rather than vague targets perhaps based on the ‘Teachers’ standards’ (DfE, 2021), which often lack the explanation of how to achieve this. This can be achieved by instructional coaching. Instead of ‘make sure you have clear rules and routines in the classroom’ (TS: 7a), the instructional coach guides a new teacher on the ‘how’ by explaining and modelling – for example, by employing a countdown strategy to instil readiness at the beginning of a task. This approach draws on the ‘borrowing and reorganising principle’ of CLT, which states that most of the information in our long-term memory is obtained from other individuals (Sweller et al., 2019). Instructional coaching can help to reduce element interactivity, not only by increasing expertise but also by simply reducing the number of elements needing to be managed.

When I needed to focus on classroom practice, an instructional coach could have provided pre-planned lessons. When I needed to develop my lesson planning expertise, they could have provided partially completed plans, modelled their own planning, thinking aloud, and engaged in discussions on how best to approach lesson planning and use of data. This would reduce my element interactivity by reducing the number of actual elements in play, simultaneously increasing my expertise though the ‘borrowing and reorganising principle’, reducing element interactivity even further. Instructional coaching programmes are directive in their approach to teacher improvement because they address the specific skills that new teachers need to develop. New teachers need to increase their expertise as rapidly as possible so that they are equipped to deal with the challenges and complexities of the teaching profession. However, it is very important to be aware that the approaches suggested in this article become less suitable as teacher expertise increases. Berliner suggests that it takes somewhere between five and seven years to become an ‘expert’ teacher, but also notes that experience alone will not make a teacher an expert; ongoing development requires dedicated work (Berliner, 2004). However, the ‘expertise reversal effect’ explains that although increasing expertise decreases element interactivity due to automaticity, ‘instructional procedures designed for novices dealing with multiple, interacting elements can be counterproductive as expertise increases’ (Sweller et al., 2019, p. 277). It has been shown that taking instructional approaches for more expert teachers can eventually increase element interactivity (Kalyuga, 2007). As teacher expertise increases, they need to be given choice and flexibility in their ongoing development (Wiliam, 2016). However, this is a discussion for another time.