How many times have you trained a junior colleague, new hire, or summer intern in a task only to have that person come knocking on your door every five minutes with a different question about some key detail?
Let me ask you a different question. Do you remember sitting in your physics, or chemistry, or calculus class in high school or college and watching the teacher do a problem on the board? Do you remember being able to solve a problem on your own if the new problem was just like the old one? Do you remember not being able to solve a new problem if it was not just like the old one?
One reason this occurs is because your teacher was (presumably) a subject matter expert and one of the ironies of being an expert is that you often lose touch with what it is like to be a novice. When your new hire has a raft of questions, even after your careful training, it could be because you are now an expert and are falling into the same trap. Part of becoming an expert, after all, is that certain aspects of problem-solving just become automatic. Experts often are unable to articulate the many "obvious" (to them) things they do when carrying out a procedure or solving a problem. This is true whether the expert is a high school math teacher, or a manager trying to train a new hire. These experts can articulate what they think they do, but they do not, and perhaps cannot, articulate what they really do.
To address this issue, a great many people have developed any number of techniques and products aimed at improving training and instruction. But focusing on the technique is putting the cart before the horse.
First, the focus must be on identifying what a learner needs to know.
There are a variety of methods for identifying the knowledge needed for carrying out tasks and solving problems; many of these techniques fall under the heading of "task analysis." Unfortunately, these techniques tend to be formal and can be cumbersome to use or require some background in cognitive psychology. Plus, these techniques tend to allow the expert to say what he or she does without necessarily holding their feet to the fire! I believe the best way to identify what experts do is to have them solve the problems or carry out the tasks in question and to require the expert to justify the steps he is taking as he takes them.
This tight coupling between steps and "theory" is an important feature of the task analysis technique I have developed for use in my problem solving and instructional design research and consulting. If experts were given free rein to describe problem solving procedures in their domain, they often would lean towards "big picture" theory that would not necessarily connect with anything the learner understands and, consequently, would not be particularly meaningful or useful to the learner. A tight coupling of theory, as needed, with steps, goes a long way towards guiding the construction of instructional and training materials that will be much more useful for the learner.
For instance, I have been working with experienced college physics instructors to develop better instructional materials for students. I asked them to create a set of problems — representing a part of the course — that a student in introductory physics should be able to solve if he or she "understood" that part of the course. As these instructors solved the problems while talking out loud, I required them to justify or explain every step while I furiously took notes. One particularly striking result of this process was how often the instructors had to stop and scratch their heads as they tried to provide a justification for their steps. The justifications in this case were rooted in laws of physics, but the relevant features or implications of the laws were things that the instructors had internalized or automated and they struggled to make them explicit to me. The resulting solutions — after much back and forth — were of course quite lengthy and verbose, but they ultimately provide the raw material for guiding the construction of better worked examples and lecture materials for learners. I use the notes to then solve new problems with the expert available to help me when I don't know what to do. Each time I reach an "impasse" I revise the notes. Ultimately, I reach a point where I can solve all problems the expert gives me.
I have used this approach in training projects ranging from ballet instruction to learning about computer algorithms. The common theme is to have someone who is not a subject matter expert (SME) work closely with the SME to help them identify what they really do. The non-expert must pepper the expert with questions as she works through the task to make explicit her every step, substep, and assumption.
In other words, the best way to start to train a novice in any field or to develop good instructional materials is for the expert to actually do the tasks in question. There is just no substitute.
Richard Catrambone is a Professor in the School of Psychology at the Georgia Institute of Technology. He has consulted on instructional design and human-computer interaction topics for various organizations. His teaching and mentoring has been recognized by multiple teaching awards.
Thanks to Richard Catrambone / Blogs HRB / Harvard Business School Publishing