FACTORS AFFECTING NORMING: A DEVELOPMENTAL STUDY OF ORAL LANGUAGE MEASURES IN SPANISH-SPEAKING ENGLISH LANGUAGE LEARNERS

Abstract

The Latino population of which English Language Learners (ELLs) is a subset, has demonstrated substantial growth in recent years (U.S Census Bureau, 2008), highlighting the need for normative information regarding their language skills. However, requisite to obtaining normative information is determining appropriate norming methods. The principal purpose of the present study was to ascertain appropriate norming procedures for the language variables: Mean Length of Utterance (MLU), Number of Different Words (NDW) and Words Per Minute (WPM) in English and Spanish narratives of Spanish-speaking ELLs. The issues were 1) whether age or grade norms should be used as an index of language development, 2) whether cross-sectional or longitudinal data should be utilized, and 3) whether the inclusion or exclusion of children with missing data or grade repeats affects the language measures. It was hypothesized that due to the syntactic and lexical differences that are present across languages, there would be a different developmental schedule of development for the language variables in the English and Spanish of ELLs. Participants were typically developing kindergarten to second grade Spanish speaking ELLs enrolled in transitional bilingual programs. A total of 605 children comprised the cross-sectional dataset and a total of 679 children were included in the longitudinal dataset. From these initial datasets, additional datasets were created to provide separate age and grade groups (for the cross-sectional and longitudinal datasets) as well as three different longitudinal datasets. Narratives in English and Spanish were elicited from each child using a story retell procedure. Analyses were carried out using Multivariate Analysis of Variance (MANOVA), Univariate Analysis of Variance (ANOVA) and Repeated Measures Analysis of Variance procedures. Results of both cross-sectional and longitudinal analyses indicated that age and grade are comparable indices of time for studying MLU, NDW, and WPM. Results also indicated that longitudinal data is superior to cross-sectional data for examining the language variables and that including or excluding subjects with missing data or grade repeaters does not significantly affect MLU, NDW, and WPM scores. Additionally, results confirm the findings from the research literature that MLU, NDW, and WPM are valid variables for studying narrative development.