Once you’ve identified the concepts that you want to teach (see An Easier Way To Define Training Goals) you have a clear plan of what needs to be taught. The next question becomes “How do you teach each concept?” There are several principles that we can apply that ensures mastery every time.
First, let’s explore what it means to know something. Whatever it is that you learned, let’s say how to calculate the hypotenuse of a right triangle, how to code, how to speak a language, or how to swim, there are concepts that you need to understand at an abstract or meta-level so that you can apply them in many situations. With a triangle, for example, you should be able to calculate the hypotenuse for any right triangle. In coding, you should be able to apply some code, let’s say an if-then statement, to many different types of conditions. In a language, you should be able to apply proper sentence structure in any sentence. In swimming, you would need to understand the principle of streamlining so that you can perform the proper form when you swim.
In all of these cases, you have achieved a level of understanding of the concepts that supersedes any one particular example of its application. In other words, you can generalize the concept to many different instances or situations. If the knowledge is generalized, then the individual can apply that knowledge to any circumstance.
Generalizations are definitions. If you inherently understand a definition, then you have internalized that concept.
For example, here is the definition for the word cup: “A cup is an open container, usually having a handle, and used chiefly as a receptacle from which to drink.”
From this definition, you can imagine many different cups. it might be a coffee mug, an espresso cup, a chalice, or even the Stanley cup. These are all different examples of cups. What is important for mastery is that no matter what you are presented with, you will be able to identify a cup. You should not confuse it with a bowl. Here is a definition for bowl: “A bowl is a rather deep, round dish or basin.”
The following figure illustrates how a bowl is derived from a cup. In other words, if you understand what a bowl is then you can understand what a cup is since it is a bowl with a handle. There are other differences, but we will leave it at that for simplicity’s sake.
How do we arrive at internalizing the generalization for a concept? We develop generalization through a process of seeing many examples of a concept. Each example that we learn helps us identify the concept when we see a new example of it.
That is how machine learning works. The AI computer program learns many different examples of a concept. When it sees a new example of a concept it calculates the probability, based on the characteristics that it sees, that its observation fits the definition of a particular concept. In other words, many examples lead to the ability to generalize a concept.
In addition to examples, students need some way of knowing whether they have mastered a concept. So there is another important element to the learning process. How does the student know if they have mastered the content?
For every concept that needs to be mastered, there should be tests or experiences that allow the student to show that they have mastered the concept. If they do not pass the test, the student needs to receive very specific feedback identifying which concepts they are missing. These knowledge gaps can then be re-visited and the student can see more examples and engage in more activities to help them master the concept.
This test-feedback loop originally emerged from information theory and is used in every self-regulating system from thermostats to bi-pedal robots.
In education, it is a simple yet highly effective way to ensure that students master every concept.
In review, you can foolproof your teaching by following this simple formula. For every concept, provide definitions, examples, experiences, and feedback. Following it backward, the acronym spells F.E.E.D. which is appropriate because we are feeding the learning process.