Corrigendum to “the relationship between cognitive ability and chess skill: A comprehensive meta-analysis” [Intelligence 59 (2016) 72–83](S0160289616301593)(10.1016/j.intell.2016.08.002)

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 an average). We did not realize this until recently. Publication bias analyses as originally reported. Trim and fill analyses estimate the number of missing studies from the meta-analysis due to the suppression of the most extreme results on one side of the funnel plot. The method then imputes the effect sizes for the missing studies based on the observed data's asymmetry to create a more symmetrical funnel plot. The adjusted meta-analytic mean effect size is also reported. This adjusted mean effect size is not necessarily a more valid estimate of the overall effect, but provides information about the sensitivity of the model to publication bias due to suppression. In the present case, these analyses indicated that studies yielding a larger-than-average effect size were missing from the Gf model (10 studies). By contrast, the analyses suggested that studies yielding weaker-than-average effect sizes were missing from the Gsm, Gs, and full-scale IQ models (1 study, 3 studies, and 1 study, respectively). Given that the asymmetry fell on both sides of the means across the models, there is little evidence to suggest a systematic suppression of particular effect size magnitudes. Publication bias analyses after Cheung and Chan adjustment. Trim and fill analyses estimate the number of missing studies from the meta-analysis due to the suppression of the most extreme results on one side of the funnel plot. The method then imputes the effect sizes for the missing studies based on the observed data's asymmetry to create a more symmetrical funnel plot. The adjusted meta-analytic mean effect size is also reported. This adjusted mean effect size is not necessarily a more valid estimate of the overall effect, but provides information about the sensitivity of the model to publication bias due to suppression. In the present case, these analyses indicated that studies yielding a larger-than-average effect size were missing from the Gf model (1 study) and Gc model (3 studies). By contrast, the analyses suggested that studies yielding weaker-than-average effect sizes were missing from the Gsm, Gs, and full-scale IQ models (1 study, 2 studies, and 1 study, respectively). Given that the asymmetry fell on both sides of the means across the models, there is little evidence to suggest a systematic suppression of particular effect size magnitudes. The overall conclusion that cognitive ability contributes meaningfully to individual differences in chess skill is unchanged; most important, the meta-analytic average of correlations between chess skill and broad cognitive abilities is similar to the originally reported value and still statistically significant (0.24, p <.001, in the original analyses, vs. 0.22, p <.001, in the corrected analyses). However, as shown below in Table 1, there are changes in some specific conclusions. Most notably, while the correlations of chess skill with fluid intelligence (Gf) and short-term/working memory (Gsm) are unaffected, the correlations of chess skill with crystallized intelligence (Gc) and processing speed (Gs) are no longer statistically significant. See Table 1 for a complete list of our reported results compared with the results using Cheung and Chan's approach, and see Fig. 1-9 for updated funnel plots. Questions can be directed to Alexander P. Burgoyne at [email protected].