In the original article two values from the studies by Chacko et al (2013) and Grunewaldt et al (2013) were incorrectly coded. With the correct coding the overall effect size of the working memory training program on inattention in daily life is SMD= -0.37, 95% CI -0.63 to -0.11. As previously reported, this overall effect is significant (p=.005). The forest plot can be viewed here: https://www.dropbox.com/s.... There were no significant differences between any of the subgroups, as previously reported.

This correction was submitted to PLOS ONE on 14 April 2015.

In comment to the analysis by Dovis, van Retergem and Huisenga, the method of introducing imputed values is always questionable. In this case it is incorrect, since the heterogeneity was not significant.


Megan Spencer-Smith and Torkel Klingberg