Cognitive Load and Psychological Safety

What does Cognitive Load have to do with Psychological Safety?

Psychological safety helps to reduce our cognitive load by allowing us to ask for help and worry less about the interpersonal risks we take. Conversely, high cognitive load can mean we’re more likely to default to silence. When our cognitive load is high, we have less available capacity to make the constant and rapid mental cost/benefit calculations of whether it’s safe to and worth it to speak up.

A model of Cognitive Load

Why are some tasks more “difficult” than others? Some tasks require us to know a lot already, such as how to code, diagnose an illness, or repair a car. For other tasks, we need to go and find out information that we don’t already know – perhaps by consulting instruction manuals or asking someone for advice. 

All tasks though, to greater or lesser degrees, require us to think and solve a problem. This might be ‘How should we approach this?’, ‘What do these symptoms indicate?’ or ‘How can I integrate this feature into our system?’

In this issue, we’re diving into cognitive load, and how it depends on, influences, and interacts with psychological safety.

Cognitive Load Theory (CLT) was created in 1988 by John Sweller, Emeritus Professor at the University of South Wales, Sydney. It suggests that our working memory, the part of our brains that gives us the capacity to learn and solve problems, is only able to hold a small amount of information at any one time. Therefore, to maximise learning, instructional methods should avoid overloading it.

Back in 2018, I came across this model in a paper about cognitive load in debriefing simulations and I’ve reimagined it here:

Adapted from Fraser, K. et al. (2018) “Cognitive Load Theory for debriefing simulations: implications for faculty development”, Advances in Simulation, 3(1). doi: 10.1186/s41077-018-0086-1.

There may be other models of cognition that more accurately reflect the complex nature of the brain, but I find this particular model really useful. What it demonstrates is the limited capacity our brains have available at any given time to solve problems. It groups the things that take up cognitive space into three categories:

1. Intrinsic cognition is all the stuff you already know.
2. Extraneous cognition is all the external stuff that you need to find out, understand, or deal with.
3. Germane cognition is active learning and problem solving. That’s the stuff of value – it’s through this that we actually solve problems in the real world and achieve things. It’s also the process by which learning is transferred from short-term to long-term memory.

Let’s take as an example, the problem of driving a car to somewhere we’ve not been before. Our intrinsic cognition includes our driving skills, knowledge of what the pedals do and how to steer and brake. Our extraneous cognition has to take in new information such as maps, signs or instructions from a sat nav. It may also have to deal with distractions, such as our passengers having an argument with each other. Our germane cognition is the space left to actually solve the problem of how to get to the new place. 

What we want to do is optimise the space available for germane cognition. A number of strategies can help here. In our navigational challenge example, these may include:

• Making navigational information easier to access, such as through using a sat nav rather than trying to read the map yourself whilst driving. 
• Reducing distractions such as turning off the radio.
• Making the car easier to drive, such as through driving aid systems like cruise control or even choosing an autonomous vehicle.
• Asking someone to help by reading out directions or keeping an eye out for signs.

Cognitive load and psychological safety

This last point has particular relevance to psychological safety. One significant way to reduce cognitive load is to ask for help, and we can’t do that if we don’t feel psychologically safe. 

Even if we don’t need to ask for help, knowing we could, through being in a psychologically safe environment, could help reduce any anxiety we feel at the prospect of the task. This will itself help to maximise space for germane cognition and learning. In Amy Edmondson’s 1999 paper, she makes the point that we act in ways that inhibit learning when we face the potential for threat or embarrassment. When we feel psychologically safe, we don’t need to perform the mental risk calculations and assess our vulnerability whenever we want to ask a question. Those constant tiny calculations are cognitively burdensome. Not only can an absence of psychological safety prevent us asking the questions that could help us solve a problem, but it reduces our cognitive capacity too, in a kind of psychological double-whammy.

This also connects with Mihaly Csikszentmihalyi’s concept of “Flow” – a state of concentration or complete absorption with the activity at hand and the situation. By freeing up our cognitive capacity to focus on the task of germane problem solving, we’re maximising our potential to reach a Flow State. 

The mental calculus of speaking up

Amy Edmondson refers to a ‘tacit calculus we engage in at certain behavioural decisions points in which we “assess the interpersonal risk associated with at given behaviour”. We weigh the potential action against the particular interpersonal climate, as in ‘if I do X here, will I be hurt, embarrassed or criticised?‘ If our cognitive load is already high, through workload, stress, environment or depth of challenge, we possess less capacity to make those calculations, and thus, we are likely to default to silence.

Cognitive load in teams

We can expand the idea of cognitive load, as Matthew Skelton and Manuel Pais do in Team Topologies, to the team level. At the team level, cognitive load is essentially the sum of all team members and the scale of the system or domain that they’re responsible for solving problems within. This is partly why teams grow in size over time – because the system they’re working on grows, and we usually deem it necessary to add more people to the team in order to cope. However, (as we’ve seen in the issue on team size) – adding more people to a team actually adds to cognitive load as well! So a better approach is to decompose the system as it grows, employing multiple smaller teams rather than one large system maintained by one large team.

High performing teams are constantly learning – whether from each other, through active learning, or from mistakes. It is through utilising the germane working memory that learning is optimised. To maximise learning, and maximise performance, it is absolutely necessary to foster psychological safety and optimise cognitive load. The responsibility of anyone on a team includes fostering psychological safety as well as reducing the cognitive load for others on the team (and people in other teams too!).

Note: Understanding the nature of Germane Cognitive Load seems to remain the “holy grail” in CLT research.

On a recent Agile Engineering Podcast we discussed cognitive load and how it relates to psychological safety. Cognitive load is also a fundamental aspect of Team Topologies, affects the scale of the domain that a team works with.