Prior research has demonstrated that bilinguals outperform monolinguals on tasks requiring set shifting – that is, the ability to shift between different ways of thinking about an object or situation. For example, bilingual children have been shown to outperform monolingual controls on false-belief tasks and on the Dimensional Change Card Sort (DCCS) task. The present study aimed (a) to examine whether the bilingual advantage in false-belief and DCCS tasks can be replicated when accounting for languages/cultures and socio-economic status and (b) to determine whether inhibitory control, metarepresentation, and/or working memory underlie the advantage, if any exists.
Three language groups (24 English monolingual, 24 French monolingual, and 24 English-French bilingual preschoolers) were tested on the following tasks: false-belief (FB) tasks, the DCCS task, an inhibitory control task (Stroop task), a metarepresentation task (Identity Statements task), a working memory task (Backward Word Span), and receptive language proficiency tests. Socio-economic status was measured through a parental questionnaire containing questions about parental income and education.
Results showed that the three language groups were equivalent on socio-economic measures. Despite having significantly lower language proficiency scores, bilinguals’ raw scores on FB and DCCS tasks did not differ from monolinguals’ raw scores. After statistically controlling for language proficiency and age, bilinguals had significantly higher FB scores, but did not differ from monolinguals on DCCS scores. Analyses were then performed to determine whether inhibitory control, metarepresentation, and/or working memory help bilinguals “do more” in FB “with less” language proficiency. Working memory emerged as the likely candidate that compensates for the negative effect of bilingual children’s low language proficiency on FB performance because, after controlling for age and language proficiency, it was the only cognitive ability that fulfilled both criteria: (a) its measure correlated significantly with FB and (b) there was a bilingual advantage over both monolingual groups in the measure. A mediation analysis confirmed that the working memory measure significantly mediates the relation between bilingual status and FB while controlling for age and language proficiency. Both components of the working memory measure – that is, understanding of task instructions and maintenance/manipulation capacity – mediate this relation.
| Identifer | oai:union.ndltd.org:TORONTO/oai:tspace.library.utoronto.ca:1807/31882 |
| Date | 11 January 2012 |
| Creators | Nguyen, Thien-Kim |
| Contributors | Astington, Janet |
| Source Sets | University of Toronto |
| Language | en_ca |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0075 seconds