“The Role of Deliberate Practice in Expert Performance: Revisiting Et Al 1993”, 2019-08-21 ():
We sought to replicate et al 1993’s seminal study on deliberate practice. Ericsson et al found that differences in retrospective estimates of accumulated amounts of deliberate practice corresponded to each skill level of student violinists. They concluded, ‘individual differences in ultimate performance can largely be accounted for by differential amounts of past and current levels of practice’ (p. 392).
We reproduced the methodology with notable exceptions, namely (1) employing a double-blind procedure, (2) conducting analyses better suited to the study design, and (3) testing previously unanswered questions about teacher-designed practice—that is, we examined the way et al 1993 operationalized deliberate practice (practice alone), and their theoretical but previously unmeasured definition of deliberate practice (teacher-designed practice), and compared them.
We did not replicate the core finding that accumulated amounts of deliberate practice corresponded to each skill level. Overall, the size of the effect was substantial, but considerably smaller than the original study’s effect size. Teacher-designed practice was perceived as less relevant to improving performance on the violin than practice alone. Further, amount of teacher-designed practice did not account for more variance in performance than amount of practice alone.
…Further, the size of the effect did not replicate. et al 1993’s comparison of practice alone between the best and good violinists combined as a single group and the less accomplished violinists explained 48% of the variance in performance. Our comparison of practice alone among the 3 groups explained 26% of the variance, which is similar to 23%, the meta-analytic average amount of performance variance explained by deliberate practice in the music domain [6]. To be clear, explaining 26% of performance variance is not an inconsequential amount. However, this amount does not support the claim that performance levels can ‘largely be accounted for by differential amounts of past and current levels of practice’ (p. 392, emphasis added).
Implications for the deliberate practice theory are discussed.