Sentence Repetition in Typical and Atypical Spanish-Speaking Preschoolers Who Are English Language Learners

Abstract

Limitations in phonological working memory are frequent in children with primary language impairment (PLI) across several languages. Nonword repetition and sentence repetition have been evaluated in bilingual children exposed to a variety of language pairs. Nonword repetition tasks assist in the identification of bilingual Spanish-English children with language disorders when both languages are considered. This study investigates the clinical diagnostic accuracy of Spanish and English sentence repetition tasks for Latino Spanish-speaking preschoolers who are English language learners. Associations between language ability and phonological working memory measures are also explored. There were 40 Latino preschoolers with PLI and 40 age-matched peers with typical language development. Parent questionnaires and class observations were used to establish language use and proficiency. Affected children met criteria based on parent and/or teacher concern, and performance on the semantics and morphosyntax subtests of the Bilingual English Spanish Assessment (BESA) and a Spanish nonword repetition task. Spanish and English vocabulary was also measured. All children had nonverbal cognitive development within the normal range and no evidence of other special needs. Both Spanish and English sentence repetition tasks (SSRT and ESRT) had 21 target sentences of equal syllable length. Children repeated sentences immediately after the evaluator's presentation. Performance was quantified using a target verb and argument use score (TVAS) and a word recall score (WMEM). The children with PLI had significantly lower scores than the control group in the two languages. Discriminant function analyses were run to determine the accuracy of the measures in identifying ELLs with PLI from typical age-matched ELLs. For Spanish, SSRT TVAS had poor sensitivity (67.5%) and good specificity (92.5%). The SSRT WMEM had fair sensitivity (80%) and good specificity (92.7%). In contrast, for the English SRT, TVAS had fair sensitivity (83.3%) but poor specificity (59%). The ESRT WMEM had fair sensitivity (89.2%) and poor specificity (59%). There were significant and positive correlations between measures of language ability (e.g., BESA morphosyntax and semantics scores) and the sentence repetition scores in both languages. Spanish nonword repetition scores were also significantly and positively correlated to Spanish and English sentence repetition scores. General linguistic knowledge and phonological working memory appear to contribute to the children's sentence repetition performance in their strong and weak languages. English vocabulary scores, on the other hand, were significantly correlated to English sentence repetition scores but not to Spanish sentence repetition scores, indicating that language-specific lexical skills play an important role in sentence repetition. In conclusion, the Spanish sentence repetition task appears to be an accurate identification measure for PLI in Latino ELLs when the word recall scoring is used. The task in English presented increased difficulty to both typical and atypical preschoolers. A minimal level of vocabulary may be necessary for English language learners to be successful in sentence repetition tasks in the weak language. These results underscore the need to consider English language proficiency level when assessing ELLs with or without language disorders. (PsycINFO Database Record (c) 2019 APA, all rights reserved)