Since the publication of their now-classic study of deliberate practice in a Berlin conservatory, Ericsson, Krampe, and Tesch-Römer’s 1993 article has been the subject of a lot of discussion. It was the key source of Malcolm Gladwell’s “10,00 Hour Rule” chapter in Outliers, and people who study skill acquisition, expertise, and genius— and who argue about how much genetics, social factors, and practice play in making people world class— almost all cite it, if only to try to demolish it.

One of the claims Ericsson et al make is that deliberate practice is the thing that determines who becomes a great violinist, and who goes on to an okay but not great career. Now, a new article on “Deliberate Practice and Performance in Music, Games, Sports, Education, and Professions” performs a meta-analysis of 88 studies of deliberate practice, in order to answer the question, “How much of the total variance in performance is explained by the accumulated amount of deliberate practice?”

The answer, they say, is that it matters— but the impact varies a lot from field to field.

Moderator analyses revealed that the strength of the relationship between deliberate practice and performance varied by domain. In terms of percentage of variance in performance explained, the effect of deliberate practice was strong for games (26%), music (21%), and sports (18%), and much weaker for education (4%) and professions (< 1% and not statistically significant). Why were the effect sizes for education and professions so much smaller? One possibility is that deliberate practice is less well defined in these domains. It could also be that in some of the studies, participants differed in amount of prestudy expertise (e.g., amount of domain knowledge before taking an academic course or accepting a job) and thus in the amount of deliberate practice they needed to achieve a given level of performance.

Moderator analyses further revealed that the effect of deliberate practice on performance tended to be larger for activities that are highly predictable (e.g., running) than for activities that are less predictable (e.g., handling an aviation emergency), as we hypothesized. Furthermore, the effect of deliberate practice on performance was stronger for studies that used retrospective methods to elicit estimates of deliberate practice than for those that used a log method. In fact, for studies using the log method, which presumably yields more valid estimates than retrospective methods do, deliberate practice accounted for only 5% of the variance in performance. This finding suggests that the use of what Ericsson (2014) termed a “high-fidelity” (p. 13) approach to assessing deliberate practice (e.g., video monitoring) might reveal that the relationship between deliberate practice and performance is weaker than the results of this meta-analysis indicate. Finally, the relationship between deliberate practice and performance was weaker for studies that used a standardized objective measure of performance (e.g., chess rating) than for studies that used group membership as the measure of performance.

For those of you who are a little geekier, here’s the statistical stuff:

Domain was a statistically significant moderator, Q(4) = 49.09, p < .001. Percentage of variance in performance explained by deliberate practice was 26% for games (Formula = .51, p < .001), 21% for music (Formula = .46, p < .001), 18% for sports (Formula = .42, p < .001), 4% for education (Formula = .21, p < .001), and less than 1% for professions (Formula = .05, p = .62; see Fig. 3).

[T]he percentage of variance in performance explained by deliberate practice was largest (24%) for activities high in predictability (Formula = .49), intermediate (12%) for activities moderate in predictability (Formula = .35), and smallest (4%) for activities low in predictability.

Personally I love the Ericsson article, but I think that music education is a bit unusual, in ways that limit the generalizability of their results. Music education is quite structured; pretty much everyone agrees on what a great performance is, what counts as a subtle interpretation, what technical skills a great performer has to have, and so on. The more well-defined your field is, the more you can have deliberate practice to prepare for it. In contrast, in lots of professions deliberate practice is do because the standards guiding good professional practice are ill-defined.

Still, I quite like the original article, and have found it quite inspiring in my current book project.