'Peak': An Interview With Anders Ericsson & Robert Pool
Anders Ericsson and Robert Pool are co-authors of the new book, Peak: Secrets From The New Science Of Expertise. They graciously agreed to answer several of my questions.
LF: Much has been written - both supportive and critical - about the role of deliberate practice in developing expertise, as well as how the so-called "10,000 hour rule" relates to it. It's seemed to me that much of the criticism about deliberate practice has missed the point - that though there may be other factors involved in developing expertise, deliberate practice is the key one that is actually entirely within a person's control. And the 10,000 hour "rule" never seemed to be one you had come up with, but had been popularized by others.
Can you give a short-and-sweet list of steps involved in deliberate practice, explain what its relationship is to factors like genetics, and quickly explain the confusion around the 10,000 hour rule?
Anders Ericsson & Robert Pool:
The sorts of activities that most people consider "practice" are generally not very helpful in improving one's performance. A golfer gets in at least eighteen holes every week and tries to hit a bucket of balls at the driving range beforehand. A pianist plays the same exercises over and over until they are completely automatic. A teacher has been teaching the same material in the same way for twenty years, thinking that practice must eventually make perfect. None of these people are likely to see much improvement despite all that "practice."
To be truly effective, practice must follow a set of principles that we understand quite well now. First, the training must be done at the edge of one's comfort zone. If it's easy or it's automatic, you're not going to be improving much. Second, effective practice needs to be purposeful. You are doing what you are doing with the specific goal of improving one particular aspect of your performance. Finally, effective practice requires feedback--you need to know what you're doing wrong so you can figure out ways to correct it. The resulting progress comes as a series of baby steps, none very impressive on its own, but they can add up to an incredible journey.
In our book we call this approach "purposeful practice." Deliberate practice is a specialized--and particularly effective--form of purposeful practice where an experienced teacher designs the training exercises and monitors a student's progress, modifying the training as necessary to keep the student progressing steadily.
Deliberate practice and other types of one-on-one training are relatively rare in sports and the arts, but are quite common in music and ballet instruction. The roots of the "10,000-hour rule" are to be found in a study that I and two colleagues did of a group of student violinists at an international music academy in Berlin. We found that the elite students in that group had spent, on average, about 10,000 hours practicing music under the instruction of teachers by the time they were 20 years old, and this was significantly more hours than the less accomplished violinists at the same music academy. Malcolm Gladwell, in his book Outliers, took this result and combined it with a few others to get what we said was a general rule--that people who were at the top of their field in pretty much any area had put in 10,000 hours of practice to reach that point. In reality there is nothing magical about 10,000 hours--winners of international piano competitions have generally spent over 20,000 hours practicing, for instance--and what should be the takeaway message of the 10,000-hour rule is that nobody reaches the highest levels in a field with objective performance standards unless he or she has spent many, many hours--not necessarily 10,000, but a lot--on purposeful and deliberate practice.
Many people believe that the top people in any field must have some genetic advantage over the rest of us. There are certainly a few genes--specifically the genes for height and body size--that can have an important influence over a person's success in some fields. A man whose genes destine him to be no taller than 5-foot-8 will have a difficult time becoming a professional basketball player. The ability to sustain concentration and adapt to training demands might be other areas where individual differences in genes could have an influence on a person's ultimate success in a field. But at present no one has found any clear and compelling evidence that having certain genes (outside the genes for height and body size) is necessary to attain elite performance in a particular field. To the extent that some genes do play a role in influencing a person's ultimate potential in a given area, it is our guess that these genes will probably act by shaping how a person responds to purposeful and deliberate practice.
LF: It was exciting to read in your book about Nobel Prize winner Carl Wieman's work applying deliberate practice to his teaching, which he described here in a guest column. You compare traditional teaching and learning in schools with his work as as "skills versus knowledge - what you can do versus what you know." Can you elaborate on what you see as the difference?
Anders Ericsson & Robert Pool:
Traditionally, education has relied in large part on lectures and reading, with students' knowledge being tested with multiple-choice questions or other sorts of questions designed to gauge how much information the students have retained. The implicit assumption in this approach is that if you provide the students with the right sorts of knowledge, you can rely on them to apply the knowledge appropriately when it comes time for them to actually do something with it. An excellent example of this in the work world is the typical approach to continuing medical education, which has doctors sitting in a large room and listening to a lecture, probably with a PowerPoint presentation--again with the assumption that the doctors can go back to their practices and use that knowledge to improve their performance.
Unfortunately, study after study has found that this approach to education is not successful in training students to think and reason and that they often fail to be able to apply their knowledge to solve problems in real world situations. In the case of doctors, which has been particularly well studied, there is good evidence that simply listening to lectures does nothing to improve a doctor's performance when that performance is judged in terms of how well the doctor's patients do.
Training in such areas as music, dance, chess, and sports takes a completely different approach. While there is inevitably a certain amount of knowledge a student in one of these areas must develop, that knowledge comes about as part of the process of developing the skills that are required to excel in that area. In particular, training in those areas is focused on the development of mental representations which are, in essence, mental tools that improve one's ability to think and reason about real-world situations and also allow a person to monitor and evaluate his or her own performance, spot mistakes, and figure out what sort of training will help eradicate those mistakes. Carl Wieman and his group had students first pick up the necessary knowledge for each lesson on their own by watching videos of lectures and reading, and then they used the classroom time for students, under the supervision of a teacher and teacher's aide, to develop their skill at solving physics problems. In other words, the focus during class was on helping the students learn to do--in this case, to solve problems in freshman physics--rather than on having them accumulate knowledge. One key result here is that the knowledge actually came as part of the package--the students knew the material as well as students taught in a traditional physics class, but they could solve the problems much better.
LF: You share motivation suggestions to help anybody, including students, be energized to apply deliberate practice. They are good ones, as far as they go. However, most assume that people are already seeking ways to motivate themselves.
What your suggestions to teachers (and to parents) who are teaching or raising children who may not be exhibiting much of a desire towards wanting to apply deliberate practice to academic improvement? How can we help young people develop intrinsic motivation to do so?
Anders Ericsson & Robert Pool:
Like anyone, young or old, children are motivated by seeing that their efforts lead to something that they find rewarding. We do not think of motivation to practice as a fixed characteristic, something that a child either has or does not have, but rather as something that develops along with a child's performance in a given domain. A young child may start learning the piano for various reasons--parental pressure, a best friend is learning, and so on--but ultimately the child will continue practicing only if he or she is getting some sort of perceived benefit from it. A young musician may like the acclaim and recognition that comes from playing the piano well, for example, or may like the sense of accomplishment and satisfaction that comes from mastering a difficult piece. Whatever the perceived benefit may be, it is important that students can see that their practice is instrumental in helping them improve and attain a higher level performance, and ideally they should be able to see that improvement on a daily or weekly basis.
One of the most important ways to help students become motivated is to encourage their ability to think and reason independently so that they can eventually take control of their own practice and performance. As long as students are following someone else's training instructions, it is difficult for them to feel motivated by anything beyond the pleasure they get from various accomplishments. They feel relatively little "ownership" of their skill. But once they can tell themselves "I did that!" about a particular accomplishment, it feels good, and they want to do it some more. Another form of motivation is the pleasure that comes from performing a particular task well, whether it is playing a piece on the piano or solving a math problem. A key step in all of this is for the the teacher or parents to help the students develop effective mental representations so they can monitor their performance and think and reason about the problems that arise in the performance.
Some specific advice: When a student is first learning a skill, it is critical to limit their practice time to 15-20 minutes per day so they are able to concentrate fully and make observable improvements. In the beginning parents will need to monitor practice so as to give their children feedback about errors and also give them enthusiastic praise for attained improvements. Over time they will be able to take control and ownership of their practice and internalize their motivation.
LF: One of your chapters offers a very critical "take" on professional development provided to physicians. Much, if not all, of it seemed to me also application to teacher professional development. Could you summarize your critique of doctor "PD" and, if possible, make an observation on how you think it can be applied to teachers?
Anders Ericsson & Robert Pool:
For a long time the training of doctors has focused on communicating knowledge, particularly in the case of medical students, and little attention was paid to the need to train skills. There was a famous saying about how doctors learned medical procedures: "See one, do one, teach one." That is, it was assumed that it was enough to see a procedure done once--an assumption that sprung from the emphasis on knowledge rather than skills. However, research has shown that the performance of surgeons--as measured by patient outcomes--improves steadily as a function of the number of completed procedures, at least through the first 50 to 150 surgical procedures. It is nowhere near enough just to know how to perform a surgery; surgery is a skill that must be developed through practice. Today, in order to minimize the negative consequences for patients from their doctors learning on the job, surgeons first train intensively with simulators and then perform their early surgeries under the supervision of an experienced surgeon. As a result, patients can expect to get excellent care from all surgeons, even those who have just graduated.
The training of teachers could be similarly improved by designing learning environments where teachers could get immediate feedback on their ability to make the classroom an effective learning environment and on how well they diagnose the problems of individual students in order to help them succeed. If teacher candidates could be provided with learning environments where they could get immediate feedback on their actions and reactions and thus learn from their mistakes without harming the education of any students, they would likely be more effective teachers once they start their professional careers.
LF: Following up the previous question, you describe that ideal deliberate practice requires mentors to observe a person in action in order to provide feedback. A number of school districts employ "experts" to observe teachers and provide high-stakes feedback as part of an evaluation process that is used to determine if the teacher will keep or lose their job. Is it safe to say that high-stakes situations like those are not the same as providing support for deliberate practice?
Anders Ericsson & Robert Pool:
You are correct that the sort of evaluation you describe, which is intended to provide an overall measure of a person's performance, is not the same as the feedback used in deliberate practice, whose purpose is to identify specific problems and weaknesses so that they can be corrected. However, there is a larger issue here related to the question of how one should evaluate performance. This is a crucial issue in the expert performance framework, where the starting point is to identify those individuals who are truly experts--that is, who have reproducibly superior performance.
How should one go about identifying expert teachers objectively? One objective measure would be to examine the improvement in the performance of a teacher's students over the course of a year. Very good teachers will improve the performance of their students by motivating them to acquire the target skills and by diagnosing challenges for students and designing appropriate practice in the classroom as well as--with the help of parents--outside the classroom. Teachers whose students improve consistently more than the students of their colleagues (after controlling for starting performance) can deservedly be called expert performers, and these teachers should be recruited as coaches and mentors for helping their colleagues.
Experience with other fields suggests that the high-performing teachers as identified by the objective measure of improvements in student performance are likely to be a different group of teachers than those who are normally considered to be "the experts." It is generally those teachers with the most teaching experience or the higher degrees in their subject area (masters and Ph.D.'s), who are judged by their colleagues to be the most expert, but we now know that experience and education are only weak predictors of whether a teacher's students will have better learning outcomes.
LF: You describe some successes in using technology in promoting deliberate practice. It seems to me that using tech in this way could also potentially lead to problems. What "do's and don't's" would you offer to people who are considering using technical tools in this way.
Anders Ericsson & Robert Pool:
The main rule should be to keep the principles of deliberate practice in mind when designing the uses of new technologies. For example, advances in technology make it possible to give all of the students in a class computers so they can simultaneously answer questions that test their understanding of the key concepts and methods presented by the teacher. In particular, this gives the teacher the chance to get immediate feedback on each student and personalize instruction based on that feedback. For example, a teacher could use computers to analyze students' responses after the end of the school day to reflect on the day's learning and to identify students who need help to master the topics. As computer-based methods are developed and improved, the teachers should be given better diagnosis of students' problems as well as recommendations for computer-administered additional computer-based training for students in need.
Technology could also be used to help teachers develop their own skills. Imagine, for instance, a set of training sessions based on video recordings of actual classes. A student teacher would watch the video up to a point, at which time the screen would go blank and the teacher candidate would be asked to predict what will happen next. With such lessons it is possible to train teachers to anticipate disruptions in the classroom and to develop behaviors to intervene before the disruptions happen. With a library of videos of various types, it should be possible to help teachers improve their skills in different areas with the same approach--having the teacher experience a number of similar events, react to each in turn, receive feedback on how to do better, then go back and try again. In one afternoon a teacher using this video library might experience and react to as many events of a particular type as another teacher might need years to experience. The same approach could be used to promote increased motivation among students as well as increased support from parents.
LF: Thanks, Anders & Robert!