Assessing Bilingual Children's Language and Cognition in Their Dominant Language

• In Event: 31.035 - The Role of Heritage Languages for Academic Achievement: International Research Perspectives

Abstract

The language landscape in the US education system has grown increasingly diverse. According to the 2011 American Community Survey (Ryan, 2013), approximately 9 million children between the ages of 5 and 19 reported speaking a non-English language at home. Given these multilingual children are primarily educated in English, one current challenge for stakeholders of education is to identify fair and sensitive assessments for children with diverse language backgrounds. The present study demonstrates the applicability of assessing Spanish-English bilingual children in their dominant language and using a non-verbal executive function task to provide a comprehensive understanding of language and cognition in Spanish-English bilingual children.
Fifty-five Spanish-English bilingual preschoolers attending government-subsidized programs participated. According to a home language survey, 23 and 32 children were from English-dominant and Spanish-dominant homes respectively. These children were given the Kaufman Brief Intelligent Test-2 (KBIT-2, Kaufman & Kaufman, 2004), a phoneme judgment task, an executive function task and three subtests (Concepts and Following Directions, Sentence Structure and Word Class-Receptive) from the Clinical Evaluation of Language Fundamentals-Preschool-2 (CELF-P2, Semel, Wiig & Secord, 2004). While the KBIT-2 was administered in English, the other tasks were administered in the children’s dominant home language.
Repeated measure analysis of variance showed a significant group x condition interaction in the KBIT-2, F(1, 52) = 29.6, p< .0001. The English-dominant and Spanish-dominant children were comparable in the nonverbal domain with the Spanish-dominant children performed significantly lower in the verbal domain. In phoneme judgment, both groups of children had better than chance judgments for phonemes general across English and Spanish and specific to Russian (a foreign language to all children), |t|s > 2.8, ps < .004. However, English-dominant children were slightly better than chance when judging English-specific phonemes, |t| = 2.5, p< .02, while the Spanish-dominant children were at chance, |t|= 1.8, ns. In the executive function task, Spanish-dominant children’s accuracy was facilitated by spatial cues whereas the English-dominant children did not benefit from the cues, F(1, 49) = 6.9, p<.02. These group comparisons showed dissociated performance in tasks with differential levels of language demand. Compared to English-dominant children, Spanish-dominant children had slightly inferior performance in tasks requiring oral language comprehension but had comparable or better performance in tasks that demanded minimal language processing.
Compared against the CELF-P2 standardized statistics (μ = 10, σ = 3), the English-dominant children had significantly lower performance in all the subtests administered in English, Zs < -2.1, ps < .04. The Spanish-dominant children had significantly lower Spanish oral language proficiency in Concepts and Following Directions and Sentence Structures, Zs < -3, ps < .001, but their performance in Word Class-Receptive was not different from the reported mean, |Z| < 1. These findings suggest that testing in children
’s dominant language reveals children’s general language proficiency despite their weaker proficiency in their second language. These findings support assessments in children’s dominant language and/or using nonverbal executive function tasks to gain a comprehensive understanding of multilingual preschoolers’ language and cognitive development.

Authors

• Gigi Luk, Harvard Graduate School of Education

• Sibylla Leon Guerrero, Harvard University

• Sarah Smith