| Speech-language assessment of bilingual Spanish-English speaking children in the United States has been difficult for researchers and clinicians because there are few assessment measures that have been comprehensively studied with this population, and none have shown to be optimal. In recent years, additional assessment measures have been recommended for this population, but few studies have examined the effectiveness of these measures to accurately differentiate children with typical language from children with language impairments. This study examined whether a specific assessment measure is best in differentiating typical language learners and children with language impairment; and whether there are a group of assessment measures that collectively best differentiate the TL and LI groups of children. Selected language assessment measures were administered to 30 Mexican-American 4 and 5 year old bilingual Spanish-English speaking children (15 TL and 15 LI) to determine the predictive value of each assessment measure, as well as which group of predictors can appropriately identify LI and TL children. Using a multiple linear regression analysis, this study found that all of the assessment measures except the dynamic assessment task and NWR English task can be used to assist in the identification of language impairment. Spanish standardized assessments were highly discriminating and strongly correlated with language status. Other assessment measures including the English standardized assessments, language sample, parent interview, and NWR Spanish tasks were highly discriminating and moderately correlated with language status. The results from this study suggest that a parent interview, Spanish and English standardized assessment, and language sample can be administered as a comprehensive battery to identify children with language impairment. |
