Bilingualism has been associated with enhanced executive function (EF) (Bialystok, Craik, Green & Gollan, 2009). Research in this area has primarily compared the performance of monolinguals and bilinguals on executive function tasks. To examine if there are differences in this “bilingual advantage” within bilinguals, Chen & Zhou (2013) examined differences in EF among English-proficient immigrant children who varied in degrees of heritage language proficiency. They found that heritage language literacy and productive vocabulary predicted differences in accuracy on a Go/No-Go task, suggesting that enhanced EF in bilinguals may be observed as a within-group variable.
The bilingual advantage has been found at various developmental stages from infancy through adulthood (Bialystok, 2011) Executive function skills, in general, develop throughout the lifespan. Results from different studies in monolinguals suggest that different components of EF may have different developmental trajectories (Best & Miller, 2010). Inhibition appears to show significant improvement particularly during the preschool years and taper off, while shifting and working memory seem to develop in a more gradual, linear manner.
We therefore wanted to examine: 1) Does degree of bilingualism predict EF performance among dual-language children of immigrants? 2) Does the relationship between bilingualism and components of EF vary across two different age groups?
The sample consisted of sixty-seven kindergartners (Mean age = 5.8 yrs; 49% male) and ninety-five 7th graders of Turkish origin (Mean age = 12.8 yrs; 59% male) residing in Norway. Measures of language and executive function were completed during a family home visit: a) Norwegian language - Norwegian version of the Expressive One Word Picture Vocabulary Test; b) Turkish language - Turkish version of the Peabody Picture Vocabulary Test – IV; c) Working Memory - Digit Span subtest of the WISC-IV; and d) Inhibition/Shift - Hearts and Flowers computer task. Scores on the two language tests were used to classify the children on a Bilingualism variable as Low/Low (dummy code = 0), Low/High or High/Low (dummy code = 1), or High/High (dummy code = 2).
A series of hierarchical regressions were carried out for both age groups. The children’s age, gender, and maternal education were entered first in the model. The bilingualism variable was subsequently entered to determine its unique contribution to the variance in EF performance. Separate regressions were run for each of the executive function scores. Preliminary analyses showed that bilingualism predicted EF performance primarily for the preschoolers, but not for the preadolescents. Among the preschoolers, bilingualism significantly predicted performance on measures of working memory (digit span forward and backward), inhibition (accuracy on the incongruent condition), and shifting/inhibition (accuracy on the mixed condition). In contrast, among the preadolescents, bilingualism significantly predicted only digit span forward, while its association with reduced mixing costs approached significance (p = .05). Results will be discussed in light of different levels of bilingualism and executive function, their developmental trajectories, and the implications for children of immigrants.