Linguistic Moderators of Effects of Definition Supports on Academic Vocabulary of Fouth-Grade Dual Language Learners

• In Event: The Role of Language Input and Production in Academic Language and Literacy Development of Dual Language Learners

Abstract

The use of monolingual and bilingual dictionary definitions to support vocabulary learning is a ubiquitous practice in English Learner (EL) instruction, yet, the field lacks experimental work that isolates and tests the effect of dictionary definitions on EL learning of general-purpose academic words.
The current study used an experimental design to study the extent to which 4th grade Spanish-speaking ELs (n=192) benefited from English or Spanish definitions during a computer-based task that required independent learning of general-purpose academic English words that were matched on grade level familiarity (6th grade according to Biemiller, 2010; Dale & O’Rourke, 1981) as well as frequency and dispersion (Zeno, Ivens, Millard, & Duvvuri, 1995). Word learning was assessed on a computer-based presentation of multiple-choice items in which students selected one of four words that best matched a definition as it appears in the Collins COBUILD English Language Dictionary (2012). Students were randomly assigned to one of three between-subject conditions in which they were presented either English definitions, Spanish Definitions, or no definitions and were asked to learn the meaning of matched cognate and noncognate target English words appearing in low or high context English sentences.
Tests of main effects revealed that students who received English definition supports were 1.56 times more likely to correctly identify the correct definition for a target English word than students in the control condition. Students who received Spanish definition supports were 1.72 times more likely to correctly identify the correct definition for a target English word than students in the control condition. The performance of students in the Spanish and English definitions conditions were not reliably different from each other.
A main effect for context revealed that students performed best when target words were presented in high-context English sentences, but the interaction between context and cognate status revealed that this effect was primarily observed for non-cognates. Results also indicate that the effect of the definitions condition was moderated by student’s degree of bilingualism as measured by performance on the Woodcock Language Proficiency–Revised (WLPR) picture-vocabulary subtest. For students who were below the sample average on both Spanish and English WLPR, the probability of a correct response decreased if they were in the Spanish definition condition and was roughly equivalent in the English definition and control conditions (see figure 1 below). For students who were at the sample average or above the sample average, the probability of a correct response was equivalent if they were in the Spanish or English definition condition yet better than in the control condition (see figure 1 below). These results suggest that the effectiveness of vocabulary learning supports may vary in relation to students combined vocabulary proficiency in English and Spanish.

Authors

• Maria S. Carlo, University of South Florida - Tampa

• Jeffrey M Williams, University of South Florida

• Mary A. Avalos, University of Miami

• Jason Anthony, University of South Florida - Tampa