Corrigendum: Deliberate Practice and Performance in Music,Games,Sports,Education,and Professions: A Meta-Analysis

Abstract

Original article: Macnamara, B. N., Hambrick, D. Z., & Oswald, F. L. (2014). Deliberate practice and performance in music, games, sports, education, and professions: A meta-analysis. Psychological Science, 25, 1608–1618. doi:10.1177/0956797614535810

In our original analysis, we adjusted for dependent performance measures using a method based on Cheung and Chan’s (2004, 2008) method. Cheung and Chan’s method adjusts the sample size to be between the sample N and the cumulative sample N and applies this to the average of the dependent effect sizes. Their adjustment formula is as follows: adjusted N = ((N – 1)/C) + 1, where C accounts for the correlation between dependent effect sizes, in addition to the overall average effect size and the number of dependent effect sizes per sample. We inadvertently used the formula as follows: adjusted N = (N – 1)/(C + 1) and then applied this formula to each individual effect size (rather than to the average).

It happens that our lower N computed for each individual effect generally offset the higher N for the cumulative effect under the Cheung and Chan approach. The formula and application we used produces values similar to those produced by robust variance estimation (another option for adjusting dependent samples). There was no practical effect on the results when we reanalyzed our findings using Cheung and Chan’s approach, and the changes had no impact whatsoever on the substance of our findings and conclusions. Table 1 provides a complete list of the originally reported results that are being changed with this Corrigendum. Questions can be directed to Brooke N. Macnamara (bnm24@case.edu).

Table 1.

Results Being Corrected

Result	Originally reported results	Corrected results using Cheung and Chan (2004, 2008) adjustment
Main model (p. 1612)
Meta-analytic average correlation between deliberate practice and performance	.35, 95% CI = [.30, .39]	.38, 95% CI = [.33, .42]
Overall variance in performance explained by deliberate practice	12%, 95% CI = [9%, 15%] (88% of variance unexplained)	14%, 95% CI = [11%, 18%] (86% of variance unexplained)
I ²	84.90	88.54
Domain as a moderator	Q(4) = 49.09, p < .001	Q(4) = 36.61, p < .001
Variance in performance explained by deliberate practice for games	26% ( $\bar{r}$ = .51, p < .001)	24% ( $\bar{r}$ = .49, p < .001)
Variance in performance explained by deliberate practice for music	21% ( $\bar{r}$ = .46, p < .001)	23% ( $\bar{r}$ = .48, p < .001)
Variance in performance explained by deliberate practice for sports	18% ( $\bar{r}$ = .42, p < .001)	20% ( $\bar{r}$ = .45, p < .001)
Variance in performance explained by deliberate practice for education	4% ( $\bar{r}$ = .21, p < .001)	5% ( $\bar{r}$ = .22, p < .001)
Variance in performance explained by deliberate practice for professions	< 1% ( $\bar{r}$ = .05, p = .62)	1% ( $\bar{r}$ = .09, p = .377)
Predictability of the task environment as a moderator	Q(1) = 20.49, b = 0.14, T² = .05, p < .001	Q(1) = 11.32, b = 0.12, T² = .05, p < .001
Variance in performance explained by deliberate practice for activities high in predictability	24% ( $\bar{r}$ = .49)	23% ( $\bar{r}$ = .48)
Variance in performance explained by deliberate practice for activities moderate in predictability	12% ( $\bar{r}$ = .35)	14% ( $\bar{r}$ = .37)
Variance in performance explained by deliberate practice for activities low in predictability	4% ( $\bar{r}$ = .21)	6% ( $\bar{r}$ = .25)
Method used to assess deliberate practice as a moderator	Q(2) = 16.19, p < .001	Q(2) = 18.18, p < .001
Variance in performance explained by deliberate practice for studies that used a questionnaire	12% ( $\bar{r}$ = .34, p < .001)	15% ( $\bar{r}$ = .38, p < .001)
Variance in performance explained by deliberate practice for studies that used a log method	5% ( $\bar{r}$ = .22, p < .001)	4% ( $\bar{r}$ = .21, p < .001)
Method used to assess performance as a moderator	Q(3) = 14.41, p = .002	Q(3) = 9.75, p = .021
Variance in performance explained by deliberate practice for studies that used laboratory tasks	14% ( $\bar{r}$ = .37, p < .001)	12% ( $\bar{r}$ = .35, p = .012)
Variance in performance explained by deliberate practice for studies that used expert ratings	9% ( $\bar{r}$ = .30, p < .001)	11% ( $\bar{r}$ = .34, p < .001)
Variance in performance explained by deliberate practice for studies that used standardized objective scores	8% ( $\bar{r}$ = .28, p < .001)	10% ( $\bar{r}$ = .32, p < .001)
Additional model excluding team sports (pp. 1612, 1615)
Overall variance in performance explained by deliberate practice	11%	12%
Variance in performance explained by deliberate practice for games	26%, p < .001	24%, p < .001
Variance in performance explained by deliberate practice for music	21%, p < .001	23%, p < .001
Variance in performance explained by deliberate practice for sports	19%, p < .001	16%, p < .001
Variance in performance explained by deliberate practice for education	4%, p < .001	5%, p < .001
Variance in performance explained by deliberate practice for professions	< 1%, p = .62	1%, p = .377
Additional model with only solitary deliberate practice (p. 1615)
Overall variance in performance explained by deliberate practice	11%	14%
Variance in performance explained by deliberate practice for games	23%, p < .001	22%, p < .001
Variance in performance explained by deliberate practice for music	23%, p < .001	25%, p < .001
Variance in performance explained by deliberate practice for sports	22%, p < .001	18%, p < .001
Additional model with only solitary practice and excluding team sports (p. 1615)
Overall variance in performance explained by deliberate practice	10%	14%
Variance in performance explained by deliberate practice for games	23%, p < .001	22%, p < .001
Variance in performance explained by deliberate practice for music	23%, p < .001	25%, p < .001
Variance in performance explained by deliberate practice for sports	28%, p < .001	21%, p < .001

Note: CI = confidence interval.

References

Cheung

S. F.

Chan

D. K.-S.

(2004). Dependent effect sizes in meta-analysis: Incorporating the degree of interdependence. Journal of Applied Psychology, 89, 780–791.

Cheung

S. F.

Chan

D. K.-S.

(2008). Dependent correlations in meta-analysis: The case of heterogeneous dependence. Educational and Psychological Measurement, 68, 760–777.