Liftology 15: Coupling models, examples, and practice

While each of the elements - models, examples, and practice - are necessary, non are sufficient in and of themselves. What matters is what you do with them: sequencing, spacing, and more. There are alternatives, but by understanding their relationships to each other, you can better choose which to use when.

So, models are abstract conceptual models of how the world works. They should effectively be causal and connected. You use them to guide predictions of outcomes of different choices, so you can evaluate those outcomes and choose the best course of action.  

However, models alone don’t lead to learning. They’re abstract, and the problems we need to solve are concrete. What evidence suggests is that to begin to understand the use of models, we need examples of how they’re used to solve problems and make decisions. Thus, we present a particular context with a problem, and show how using the model solves it. We can, of course, also show how misusing the model, such as using it in appropriate situations, not following it precisely, or using the wrong one, doesn’t solve the problem. 

Models and examples alone may seem like enough, but what really matters is using the models in practice. After we’ve seen one or a few examples, we need to then apply the model to solve problems, in particular problems like need to be solved after the learning experience. We’re developing an ability to do, and to do that you, well, need to do! Along, importantly, with feedback about what the consequences of the choices are, and if learners chose the wrong action, why it was wrong and what the right answer is. 

Presenting models, and examples, and practice provide a foundation for learning. Of course, to cement understanding, we likely need extended practice. Practice that increases in complexity, which may mean expanded models and more examples. Also, we need to extend these over time, and across contexts. As expertise develops, the feedback may be faded, as well. We can play with the order, for instance presenting a complex problem that the learner knows is relevant and needs to be solved, but can’t yet, so that the models and examples subsequently make sense. Or we can start with the models if learners already ‘get’ that the learning is needed. The correct order is what will work for the learning, but what matters is that we have all the elements clearly articulated. With that, we have a great likelihood of actually developing a new ability to ‘do’, which is (or should be) our goal.