Meta-Analysis Reveals a Bilingual Advantage That Is Dependent on Task and Age

Debate continues as to whether a bilingual advantage exists with respect to executive functioning. This report synthesised the results of 170 studies, using meta-analysis, to test whether the bilingual advantage is dependent on the task used to assess executive functioning and the age of the participants. Whether the bilingual advantage might be an artefact of publication bias was also examined. The results of the meta-analyses indicated that the bilingual advantage was both task- and age-specific. Bilinguals were significantly faster than monolinguals (Hedges' g values ranged from 0.23 to 0.34), and significantly more accurate than monolinguals (Hedges' g values ranged between 0.18 to 0.49) on four out of seven tasks. Also, an effect of age was found whereby the bilingual advantage was larger for studies comprising samples aged 50-years and over (Hedges' g = 0.49), compared to those undertaken with participants aged between18 and 29 years (Hedges' g = 0.12). The extent to which the bilingual advantage might be due to publication bias was assessed using multiple methods. These were Egger’s Test of Asymmetry, Duval and Tweedie’s Trim and Fill, Classic Fail-Safe N & PET-PEESE. Publication bias was only found when using Egger’s Test of Asymmetry and PET-PEESE method, but not when using the other methods. This review indicates that if bilingualism does enhance executive functioning, the effects are specific to at least task and age. This may arise because using multiple languages has a highly specific effect on executive functioning which is only observable in older, relative to younger, adults. The finding that publication bias was not uniformly detected across the different methods raises questions about the impact that unpublished (or undetected) studies have on meta-analyses of this literature. That is, the evidence does not unanimously indicate that the bilingual advantage can be explained with respect to publication bias.

Abstract licence: CC BY

40 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.