Paul Kirschner and I have written about complex skills before (here (2018) and here (2020), a guest blog (2020) and a chapter in our book). The main reason why we go on and on about them, is because when we talk about ‘upskilling’ and/or ‘reskilling’ in organisations, both ‘business stakeholders’ and L&D people often overlook and/or misunderstand what a skill – in particular a complex skill – really is, what it takes for people (‘learners’, ‘workers’) to develop them. As a consequence, they forget or don’t think about what it means to design learning experiences or interventions for them that are effective, efficient, and enjoyable. What I mean by that here, is that learners can and will transfer what they learned to their job and apply it in a competent and flexible way.
On 13 October, I gave a presentation at the Learning Technologies Autumn Forum about improving learning transfer when designing for complex skills. I’ll summarise it here to provide context, but the main reason I’m writing this two part blog (today’s is the first part) is because we did an exercise at the end where we discussed what it would take to drive the change required to design effectively for complex skills. I promised the participants that I’d share the outputs because I think more people could benefit from them, resulting in these two blogs.
Before moving on to what it takes to drive this type of design in organisations coming Thursday (the practice), first a summary of the parts of my presentation that anyone needs in order to understand the context when it comes to the discussion on how we can drive the required change in organisations (the theory). Let’s start with a definition of complex skills.
! Note that the work in the presentation is largely based on Van Merriënboer’s 4C/ID model and a 4C/ID course that I have partnered on with Jeroen Van Merriënboer and Jimmy Frèrejan, as well as the book ‘Ten Steps to Complex Learning’ by Jeroen Van Merriënboer and Paul A. Kirschner (2017).
What are complex skills?
Here are four criteria to help determine if we’re dealing with a complex skill:
- A complex skill requires a combination of multiple constituent skills. Note that they’re not called sub skills. This is because a complex skill is more than the sum of its parts. Let’s take driving a car – which is a complex skill – as an example. Some constituent skills for ‘driving a car’ are ‘starting your engine’, ‘using your mirrors’, using your brakes’, and ‘monitoring traffic’. To competently drive a car, you don’t just implement these and other skills as ‘sub skills’. Instead, you need to be able to combine and coordinate them as required in the situation at hand.
Note that eating a hamburger or using your phone is not a constituent skill that is part of the complex skill ‘driving a car’ 😉
- Complex skills require integrating knowledge, skills, and attitudes. For driving a car, required knowledge are ‘the rules of the road’, some of the skills I’ve already mentioned, and a required attitude is ‘the willingness to drive safely and not take (unnecessary) risks’.
- Within the context of complex skills, their constituent skills are performed differently. Some constituent skills are carried out as automatic processes, for example using your mirror or your brakes will become automatised over time. Others are carried out as controlled processes. For example, if you’re driving in a city you’ve never been in before, you need all your cognitive capacity to monitor traffic and signs, while at the same time trying to find your way. It’s normal that in such situations, we tend to lower the volume of the radio or even turn it off. These controlled processes require critical thinking, problem-solving, and decision-making.
- Complex skills take time to develop. If you can learn how to do something competently in a couple of days, it’s likely not a complex skill. Learning how to competently drive a car takes a lot of time. Just passing your theoretical and practical tests and receiving your driver’s license, doesn’t make you a competent driver. At that point, you’re meeting the minimal required standards to ‘drive safely’. And you did this when someone was in the car with you, looking over your shoulder, and unconsciously influencing your behaviour. Driving alone on a dark and stormy night with little street lighting in a strange town is quite different!
Let’s say you’ve now determined that you’re dealing with a complex skill. How do you design effectively for it?
Effective design for these complex skills
The key thing to remember when it comes to effective design for complex skills (and again, ‘effective’ in this case means that people can actually improve what they do on the job through applying the newly learned skills in a flexible manner or doing their job better) is… that we need to mirror real life from the beginning; that is use meaningful whole tasks beginning with simple tasks (NOT the same as easy ones) and proceeding to more complex ones. In short, this is because when we learn, we build cognitive schemas, which are meaningful networks connecting things like concept, scripts, and principles in our long-term memory (see examples below). Strong, well-integrated cognitive schemas facilitate retrieval and application. We build stronger and better integrated schemas if we learn the complex skill in a way that mirrors real life as much and as soon as possible. That’s it in a nutshell.
Let’s explore that in a bit more detail.
When we start learning something new, our cognitive schemas with respect to what we’re learning are limited and unstructured. When we become more ‘advanced’, the connections within our cognitive schemas grow (become both broader and deeper) and become stronger. In other words, the cognitive schemas will become ‘better formed’ and ‘better integrated’ over time, which makes retrieval of information at the right time easier. This is because, when a schema is well integrated, it counts as only ONE element when we need to use it in our working memory and this, of course, makes cognitive processing way more efficient.
The key point is: The growth and strength of the connections will be better (our cognitive schemas will be better formed and integrated) if we mirror reality and help people to ‘connect the dots’ from the very start in the learning process.
Here are six key design elements to help you to effectively design for complex skills.
- Authentic whole tasks – Always choose tasks that the learner can encounter in real life and organise them in ‘task classes’ from simple to complex (NOT easy to hard). For example, in the context of driving a car, the ‘simple’ task class could be ‘driving on a familiar and quiet road with no other traffic’ while the most complex task class could be ‘driving in an unfamiliar, very busy and chaotic environment at night in the rain’. Of course, the variables that make a task class ‘simple’ or ‘complex’ depend on the task at hand. You start by offering learners ‘whole learning tasks’ within the ‘simple task class’ and let them move on to the next level of complexity when they’re sufficiently competent within that class of tasks (again, what that means depends on the tasks and the standards that determine if someone is competent).
- Variability – This refers to slight variations within a task class (so the key is that the tasks vary slightly, but the level of complexity remains the same). In the context of driving a car, the variability could be around the speed or the surface of the road. Variability is important to train for flexibility.
- Scaffolding – This refers to guidance. When a learner is new to a task within a task class, you offer them the highest level of scaffolding (i.e., support and guidance), such as a modelling example. A modelling example explains the task at hand step by step: What to do, how to do it, and why to do it that way. The learner can study this example to ensure they fully understand before they start ‘doing’. When they’re ready, you take some scaffolding away (this is called ‘fading’) and you provided ‘guided practice’ activities. There are many ways to do this. For example, in the context of learning how to drive a car, the instructor could still do most of the steering and apply the brakes, but there could be some signs on the road that the instructor doesn’t explicitly point out to see if the learner is paying attention and making the right decision based on the sign. The lowest level of scaffolding is where the learner practices independently (no guidance, although of course there will still be feedback after completing the learning task).
- Supportive information – This type of information is to support the ‘controlled processes’ (the ones that require critical thinking, problem-solving, and decision-making). It’s about providing systematic approaches to solving a problem or carrying out a task and ‘the theory’ (complex skills require a lot of domain knowledge as without that, it will become very difficult to apply the skill competently, especially in new situations). For example, the learner can use guided simulations of various traffic situations in which they need to make quick decisions. They can study applicable theory before starting the simulation and upon completing the simulation, they can compare their decisions with the ‘best’ decisions that one can make in these situations (guided feedback).
- Procedural information – This is about rule formation (if-then) for recurrent parts of the task (this that are always the same). ‘If I want to change direction, then I use my indicator’. This type of information needs to be provided just-in-time to the learner.
- Part-task practice – This is about the recurrent parts of the task that need to be automated. For example, starting to drive on a hill needs to be practised over and over again, until it becomes automated. Part-task practice is an exception and you only design for it, if it’s either about a constituent skill that is critical (e.g., it puts lives at risk if not done well) or if it really frees up cognitive capacity to focus on the more complex aspects of a task (which is why it’s important that starting to drive on a hill or parking become automated; it allows you to anticipate situations better). If you need to design for part-task practice, you need to make sure you always teach the whole task first, so that learners understand why this part of the task needs to be automated.
It’s probably clear by now that designing this way in the workplace requires significant changes on multiple levels. On Thursday, I’ll discuss ‘the practice’, based on the discussion at the conference, where we tried to answer:
- If we want to do better and drive effective design for complex skills in the workplace, what would that take?
- What would we need to get there?
 Retrieval means moving information from long-term memory to working memory, where we become aware of it.
 A simple task has very few information elements and there are not many interactions between them. A complex task has many information elements and lots of interactions between them.