Confirmation of the importance of deliberate practice in the development of excellence
For the past ten years or so, deliberate practice has become quite well known (although many are more familiar with Malcolm Gladwell's incorrect interpretation of it as the 10,000 Hours Rule). As can be read in popular publications (such as the book Peak), deliberate practice is a form of practice that plays an important role in building excellence. But in recent years, a number of publications have appeared (such as Macnamara, 2014) that suggest that deliberate practice plays a less important role than previous research showed. Anders Ericsson, pioneer in research into deliberate practice, along with Kyle Harwell, responds to the recent criticisms in recent paper.
Classic article on deliberate practice: Ericsson et al. (1993)
In 1993, Anders Ericsson (photo), Ralf Krampe and Clemens Tesch-Römer published a classic paper entitled The Role of Deliberate Practice in the Acquisition of Expert Performance. Based on an in-depth review of previous research, the authors in the article explain what deliberate practice is and demonstrate through research the important role deliberate practice has in building excellence in music.
Deliberate practice is a form of practice in which four criteria are met:
Deliberate practice is a form of practice in which four criteria are met:
- It is individualized. This means that a qualified teacher designs exercises that take into account the current level of the student.
- It is purposeful. The exercise is specifically aimed at improving one aspect of the current functioning.
- The practitioner has a mental representation of the goal. With the help of this, the practitioner can always compare his own performance with this mental representation. This allows the practitioner to receive immediate informative feedback while practicing.
- There is a lot of repetition. The practitioner always practices the same task so that it gets closer to the goal with the help of feedback.
The article describes research at an international conservatory in Berlin. Top violinists were compared to less skilled violinists and were found to have made more use of deliberate practice. Follow-up research showed that similar effects were found in many other performance domains (Ericsson, 1996, 2003, 2007; Ericsson & Lehmann, 1996; Ericsson et al., 2018).
Gladwell's simplistic rendition of the research
Ericsson et al. (1993) has been very influential. The paper has now been cited more than 10,000 times and the concept of deliberate practice is now mentioned more than 35,000 times in scientific publications. The fame of Anders Ericsson and of deliberate practice has been greatly supported since 2008 by Malcolm Galdwell's bestseller Outliers (photo), which devoted an entire chapter to Ericsson's work. Unfortunately, Gladwell put a simplistic spin on the subject of his own that has done more harm than good. He introduced the idea that there would be a 10,000 hours rule that would require 10,000 hours to achieve peak proficiency. Galdwell did not mention the term "deliberate practice" at all.
Ericsson and his colleagues never claimed that 10,000 hours of practice would lead to top expertise. They had argued that building top expertise cannot be achieved without many years of practice. The duration of the practice is indeed important. But that there would be a fixed magical number of hours is not true. And, more importantly, how practice is performed is critical (see the four criteria for deliberate practice outlined above).
Ericsson and his colleagues never claimed that 10,000 hours of practice would lead to top expertise. They had argued that building top expertise cannot be achieved without many years of practice. The duration of the practice is indeed important. But that there would be a fixed magical number of hours is not true. And, more importantly, how practice is performed is critical (see the four criteria for deliberate practice outlined above).
New article by Ericsson & Harwell in response to critical research
In addition to much research that was in line with the findings of Ericsson et al. (1993), research also emerged that seemed to question the importance of deliberate practice. Two well-known critical articles are those of Macnamara (photo) et al. (2014) and Hambrick et al. (2014).
In a recent paper, Anders Ericsson and Kyle Harwell discuss these studies in detail. Here I briefly summarize what they write. They mainly focus on the research by Macnamara et al. (2014) because the problems mentioned occur to a lesser extent in Hambrick et al. (2014).
Macnamara et al. (2014) performed a meta-analysis to map the effect of deliberate practice. On this basis, they concluded that while deliberate practice is important for building top competence, its influence is much smaller than Ericsson and his colleagues have suggested. Macnamara et al. found in their study that deliberate practice explained only 14% of the variance in performance.
Ericsson & Harwell point out that the meta-analysis used a definition of deliberate practice that differs and is much broader than the definition proposed and researched by Ericsson et al. Because of this much broader definition, many studies have been included that do not meet the four criteria for deliberate practice. Many of the studies included in the meta-analysis do not even mention the concept of deliberate practice. It is therefore incorrect to make statements based on this meta-analysis about the relative contribution of deliberate practice in building top competence.
Macnamara et al. (2014) performed a meta-analysis to map the effect of deliberate practice. On this basis, they concluded that while deliberate practice is important for building top competence, its influence is much smaller than Ericsson and his colleagues have suggested. Macnamara et al. found in their study that deliberate practice explained only 14% of the variance in performance.
Ericsson & Harwell point out that the meta-analysis used a definition of deliberate practice that differs and is much broader than the definition proposed and researched by Ericsson et al. Because of this much broader definition, many studies have been included that do not meet the four criteria for deliberate practice. Many of the studies included in the meta-analysis do not even mention the concept of deliberate practice. It is therefore incorrect to make statements based on this meta-analysis about the relative contribution of deliberate practice in building top competence.
Four types of exercise
In their paper, Ericsson & Harwell describe four types of exercise:
- Deliberate practice: this is the form of exercise in which all four of the mentioned criteria, formulated in Ericsson et al. (1993), are met.
- Purposeful practice: This involves practicing purposefully to improve a specific activity but without access to individualized feedback and guidance from a qualified teacher.
- Structured practice: this involves structured practice in groups under the guidance of a teacher but without individualized exercises (taking into account the level of the individual.
- Naive practice: form of practice in which playful practice is done without goal, feedback and individualization.
The authors note that only a small part of the studies included in Macnamara's meta-analysis and (to a lesser extent) by Hambric concern deliberate practice. In many cases structured practice is a more appropriate label.
Objectively reproducible performance
In order to be able to research the contribution of deliberate practice, the study must not only involve deliberate practice as defined by Ericsson et al. (1993), but also reliably measured top performance. Ericsson et al. emphasize the importance of an objective way of measuring the level of the investigated achievers. They referred to the criterion of objective reproducible performance. Macnamara’s meta-analysis included multiple studies that did not show objectively measured performance. For example, several studies use a subjective assessment by 1 person (such as a coach).
Reanalysis of Macnamara et al. (2014)
Based on the principles outlined above, Ericsson & Harwell reconsidered which studies by Macnamara et al. were based on a correct definition of deliberate practice and of objectively reproducible performance. Of the original 191 effects in the study by Macnamara et al., only 14 effects remained. This figure shows step by step how this selection took place. Based on this selection, Ericsson & Harwell found that deliberate practice explained 29% of the variance in performance (and 61% after correcting for attenuation - read what that means here).
What is the role of innate talent?
In the remainder of the paper, Ericsson & Harwell discuss several other aspects that may be relevant to examining factors that are important in building top competence. They focus in particular on the possible role of genetic factors. They warn against some incorrect reasoning. Like Ericsson et al. (1993) they point out that the fact that genes influence behavior and traits does not mean that this behavior or traits are not changeable. They further explain that the fact that deliberate practice does not explain all variance in performance does not mean that innate talent explains the rest.
It is often assumed that innate abilities play an important role in the development of excellence. How much talent someone would have would determine (and limit) how far someone could get. But this is unproven. In fact, evidence seems to suggest something else, namely that something like initial abilities correlate fairly high with performance in the beginning of building an expertise, but as the expertise develops this correlation becomes smaller or even disappears completely (see for example Burgoyne et al. (2016) As an aside, the use of standardized tests is inappropriate for obtaining information about the role of genetic influences on the development of expertise, as it is unclear to what extent test scores are influenced by exercise and instruction.
Today it is possible to investigate the relationship between specific genes and performance. Studies in which this happens are called GWA studies (Gene-Wide-Association studies). Until now, on the basis of such studies, hardly any relationship can be found between individual genes and performance. It is possible that combinations of genes are related to performance. However, little is known about this and the research is difficult.
It is often assumed that innate abilities play an important role in the development of excellence. How much talent someone would have would determine (and limit) how far someone could get. But this is unproven. In fact, evidence seems to suggest something else, namely that something like initial abilities correlate fairly high with performance in the beginning of building an expertise, but as the expertise develops this correlation becomes smaller or even disappears completely (see for example Burgoyne et al. (2016) As an aside, the use of standardized tests is inappropriate for obtaining information about the role of genetic influences on the development of expertise, as it is unclear to what extent test scores are influenced by exercise and instruction.
Today it is possible to investigate the relationship between specific genes and performance. Studies in which this happens are called GWA studies (Gene-Wide-Association studies). Until now, on the basis of such studies, hardly any relationship can be found between individual genes and performance. It is possible that combinations of genes are related to performance. However, little is known about this and the research is difficult.
Discussion
That people differ in how good they are at things has probably been known since ancient times. It's no wonder we wonder why. In ancient times, statements such as "divine inspiration" were sufficient for most people. This kind of explanation is now seen by many as meaningless, primitive and unsatisfactory. Later, under the influence of people like Francis Galton, we came to believe in established genetic differences between people that determine how good we can become at something. While these kinds of statements are still influential, they also become less credible over time. It is of course clear that genes influence our behavior and our properties. But that genetic differences between individual differences between people strongly determine how good we can become at something seems increasingly unlikely, based on research.
Ericsson & Harwell have written a strong reply to Macnamara's research. They do not argue that deliberate practice is the only factor that explains top competence, nor that deliberate practice alone works to improve at something. They do show that Macnamara's research does not properly investigate the role of deliberate practice. Their recalculation confirms that deliberate practice plays a significant role in building excellence.
Ericsson & Harwell have written a strong reply to Macnamara's research. They do not argue that deliberate practice is the only factor that explains top competence, nor that deliberate practice alone works to improve at something. They do show that Macnamara's research does not properly investigate the role of deliberate practice. Their recalculation confirms that deliberate practice plays a significant role in building excellence.
Comments