Poster #15 - The Impact of Multilingualism on the Association Between Mathematical Language and Mathematical Performance

• In Event: Poster Session 01

Abstract

Growing evidence indicates that there is a strong association between children’s understanding of mathematical language and their mathematical performance (e.g., Turan & De Smedt, 2022). Most of this evidence has been restricted to children who show proficiency in only one language, usually in English (e.g., Bower et al., 2020). However, there is research to suggest that second language learners show poorer language skills compared to their monolingual peers (Peng et al., 2020). Consequently, growing up in a multilingual environment might be a risk factor for subsequent difficulties in mathematical language and mathematical ability. We therefore examined the associations between different types of mathematical language, such as quantitative (e.g., more) or spatial language (e.g., above), and different mathematical domains, including numerical competencies, measurement and geometry, in monolingual and multilingual children. Our sample consisted of 154 4-5-year-olds. It included 77 monolingual and 77 multilingual children, who were matched in terms of age and sex (Mage = 4 years 8 months and 33 females in each group). Multilingualism was defined based on an adapted version of the Q-Bex questionnaire (De Cat et al., 2022). We measured children’s quantitative and spatial language skills. We also assessed their performance in numerical competencies, measurement and geometry. In addition, we included measures of general language and spatial skill, and examined their association with children’s mathematical language and mathematical abilities. MANCOVA analyses controlling for socio-economic status (SES) indicated that monolingual children outperformed their multilingual peers on quantitative and spatial language, measurement and geometry, and general language tasks (all p’s < .05; Table 1). The two groups did not differ in their numerical competencies and in their spatial skills. Path analyses (Figure 1) for the two groups revealed different paths through which distinct domains of mathematics can be predicted on the basis of different types of mathematical language. Specifically, for monolingual children quantitative language was associated with the three domains of mathematics, whereas spatial language was only associated with children’s geometry. For multilinguals, quantitative language was associated with children’s geometry and measurement, whereas spatial language was only associated with children’s measurement. Finally, we conducted a within-sample analysis of multilingual children based on their mastery level of the language of preschool instruction (i.e., Dutch). We divided the sample into weak- (n = 32) and strong- multilinguals (n = 45), and compared them on all variables under study. MANCOVA analyses controlling for children’s SES-levels indicated that strong-multilinguals were better than weak-multilinguals in spatial language and quantitative language - the two groups did not differ in the remaining variables under study. These findings indicate that during early childhood multilingual children lag behind their monolingual peers in mathematical language, and in some but not all mathematical abilities. Importantly, it seems that the associations between components of mathematical language and domains of mathematics differ based on the child’s linguistic background. Monolingual children are displaying more of the theoretically expected associations across components of mathematical language, while these are not so clearly observed in multilingual children.

Authors

• Eylül Turan, Katholieke Universiteit Leuven (KU Leuven)
  Presenting Author

• Bert De Smedt, University of Leuven