Wednesday, June 1, 2016

Should we teach simple things first?

In my time volunteering and working for Iridescent, I came to understand a certain model for how to teach design processes. Having read the Common Core and Next Generation Science Standards, I realized that the Iridescent model of teaching is unique from the more typical and common sense teaching models, but that difference is really powerful. I'd call this model core-first teaching, and I believe it's particularly applicable to teaching topics like engineering design, coding, or the scientific method.

The Core-first teaching model

So what is core-first teaching? In this model, you identify the “core thing,” which might be a process you need to do or an idea you need to understand- we'll assume it's a process here. Then you identify all of the tangential components to the core process. Now, it's not that the tangential components are not important- they are. But if all you are doing is the tangential parts of the process, then you are not truly doing the process. You have to be doing the core process to be doing this thing.

What is the main implication of this? At the youngest grades, or when students first encounter a process, the only thing you should be teaching them is the core process, and nothing else. Thus this is a core-first teaching philosophy. To use an example, in an Iridescent design challenge, the core process is building something. The first time someone shows up to an Iridescent session, you should make sure they are building.

Now as is implied by core-first, the tangential components are taught next. Each tangential component is more or less related to the core process, and should be taught from the core outward. To go back to our example of an Iridescent design challenge, testing your design is a pretty core process too, and should be taught soon after building happens. Reflection upon and iterating on building is even less core, but next. Iterating multiple times, at larger levels (i.e. starting completely from scratch, rather than redesigning one piece) is even less core. Then there are things like connecting your reflection and iterating to content knowledge, and planning your design, which are even less core. Finally, the least core practice is defining your own questions and problems to be solved, and understanding the constraints of the design. This ordering of most to least core is based on my own experience teaching engineering design to kids.

An image of Iridescent's engineering design process. Notice that the core practice of "build" is not the first step in the process, and that a core-first teaching model assumes that skills are not taught in the order that they are used by an expert.

What’s important to note is that the least core practices are often the most crucial ones to the ultimate quality of a design. For instance, planning your design is vitally important to thinking through design ideas and starting your design with a good foundation. One might say that we are setting students up for failure, by asking them to build before teaching them the parts of the building process that most improve their design. But there's a problem with that argument- the tangential processes are only abstract ideas until you understand the core process. Doing the core experience gives you the embodied knowledge to really understand how to use the tangential processes to their fullest. For example, effective planning can only occur after an intuitive knowledge is built up regarding building, i.e. after the core practice is well formed. How you can possible make a plan to build a joint before you know how to assemble a joint? Without knowing how to build, a plan is an abstract puzzle-solving exercise, rather than a tangible, implementable solution that you can build.

I've seen families make plans on their first trip to Iridescent. The plans are often nice drawings, but they are superficial at best- they often gloss over the most important details of the design, in favor of focusing on some of the less important aspects of the design. It's similar to what Zach Gage describes here as the difference between how experts and novices approach solving problems. Of course the argument can be made that the plans are poor because the families had poor planning skills, but that argument assumes that planning is a skill in it's own, rather that a subcomponent of the general core skill of building. If planning is a subcomponent of building, then it's really only useful to plan when it helps you build better, which would occur after you build intuitive knowledge about building.

Of course a consequence of this method is that when the core practice is started, it can only be done poorly. You won’t be doing the tangential practices that ensure quality. And this is absolutely true, some level of initial failure is required in a core-first teaching philosophy. Having done this process many times, I can say with certainty that initial failure is not a roadblock to eventual success, and in fact with the right scaffolding can actual facilitate that success. But this is a significant change from our current high-stakes schooling system, in which students success is expected to be immediate, and we assume that we should only teach something to students that they can be immediately and effectively successful at.

The Simple-first teaching model

So let's describe another process, which we will call simple-first teaching. In this model, you start students off with relatively low level, simple tasks, and slowly build them towards more complex and challenging tasks. Simple tasks are the easiest to do, and therefore should be the best for a beginner to engage in.

Let's take an example from the NGSS which also describes the process of engineering. In NGSS, the engineering process involves Developing and Using Models. That's potentially a really tough task- what is an easy task related to Developing and Using Models? How about "Distinguish between a model and the actual object, process and/or events the models represent." This is a K-2 task described in Appendix F of the standards. Identifying and labelling the components of a model is a pretty straightforward, low level task that you can expect all students to be successful in.

The contrast with the core-first process is clear- labelling models is at best a pretty tangential task to the process of using models, but it's simple. Children can solve the labelling task in an abstract sense. But would knowing how to label models make children ultimately better at using models? If not, then this is an easy-but-not-essential task, that looks good from a scaffolding point of view but isn't effective at fostering learning. Let's also look at the reverse process- if students have built models, will that make them more successful at identifying models? Now we can answer of course it would, and in addition they wouldn't be answering this labelling question in an abstract sense but could answer it from embodied experiences with using models.

This isn't a carefully chosen example, as the simple-first philosophy seems deeply baked into the NGSS. In an earlier bullet point in the same appendix, the NGSS standards say that "Practices grow in complexity and sophistication across the grades," which clearly implies that we should start simple in early grades. And this point is driven home even further in the point before it: "Students in grades K-12 should engage in all eight practices over each grade bad." In other words, do every part of the practice in a simple way first (simple-first teaching), rather than doing only some of the practices, the core parts, first (core-first teaching).

To be fair to the NGSS (which I generally like), they do have another standard at the K-2 level: "Develop and/or use a model to represent amount, relationships, relative scale, and/or patterns in the natural and designed world." That is definitely the core process at work, if in the most simplistic sense possible. But it worries me that the core task is listed alongside the simple but extremely-tangential task, as if both are equally important to teach at this age. Simple does allow for initial success, but does it allow for greater eventual learning?

Novelty vs. Complexity

Let me offer one more point for comparing these two methods. There are two different ways to introduce difficulty to an activity: you can add greater complexity or greater novelty. More complex tasks are basically the same kind of tasks, but more onorous to solve. On the other hand, novel tasks are made more difficult by being different from what you have solved before. Creating challenge through complexity often makes for a much less powerful learning experience than introducing challenge through novelty, which is why novel challenges are very present in engaging games. For a great discussion of this distinction, refer again to Zach Gage's explanation.

Adding complexity really just takes more effort, but doesn't add more cognitive load. It is what is called routine problem solving. We know that the main difference between novice and experts in a domain is not their ability to handle greater complexity, but their ability to handle novelty. Experts know how to apply existing schema to new problems, whereas novices really are only able to solve routine problems.

But as Zach Gage explains, we can help novices become experts, but only through the right scaffolding. We need to help them recognize structures in new domains. There's a clear parallel towards our two teaching models here. In simple-first teaching, you make a task so simple anyone can solve it, and then make it more complex at higher grades. It is of course still a routine problem- you are not pushing novices to become experts, you are just pushing them to become harder-working novices. On the other hand, in core-first teaching, whenever you add a new tangential component, you add something novel to the process. Students have to learn to incorporate and use this novel new tool in the arsenal, and if they do so, they improve their ability to do the core process better. By helping them see the multiple uses and connections between their knowledge, and not simply asking them to practice their knowledge in more onorous ways, the core-first teaching methodology is more geared towards producing experts.

What is also worth noting is that each level of integration of tangential components involves doing new, or novel tasks, not simply doing the same tasks at higher levels of complexity. Core-first teaching constantly challenges learning with novelty throughout the learning process, and for that reason (and many others), it is my go-to process for teaching any kind of design skill.

No comments: