Impact of Instruction on Connective Words for Emergent Bilingual Adolescents' Word Knowledge and Comprehension

• In Event: The Multidimensionality of Vocabulary Knowledge: Implications for Assessment and Instruction With Multilingual Students

Abstract

Objectives: The purpose of this study was to examine emergent bilingual (EB) high school students’ knowledge of different types of connective words and to investigate whether explicit instruction in connectives impacts high school multilingual students’ reading comprehension. The research questions were: (1) Do different types of connectives pose different levels of challenge for EB adolescents, and (2) Does explicit instruction in connectives impact text comprehension?

Theoretical Framework: This work was primarily informed by the Core Academic Language Skills framework (CALS; Uccelli, et al., 2015), which identifies the core skills readers need to engage with and comprehend academic texts. These skills relate to specific features of academic language, such as: dense noun phrases and nominalizations, connective words and phrases, anaphoric references that carry themes through a text, and epistemic stance markers (Uccelli et al., 2015). While multiple processes are at play in reading comprehension in adolescents (Barnes, 2015), connective words were the focus of the current research for two reasons. First, connectives reveal specific relationships between ideas in a text and are thereby essential in building a situation model of a text (eg. Kintsch, 1998). Second, connectives can be particularly challenging for EB learners, and, furthermore, different types of connectives can pose greater challenges than others (Crosson, Lesaux, & Martiniello, 2008).

Methods: EB high school students (N = 60) from five sections of advanced English Language Development (ELD) classrooms participated in this study. All students participated in all components of an intervention designed to support their reading comprehension. Using a quasi-experimental design, nine informational texts were randomly assigned to three different instructional conditions: control (no instruction), background knowledge scaffolding (BK), background knowledge + connectives scaffolding (BK+C).

Data Sources and Analyses: Students read each article (following instruction for the BK and BK+C conditions) and completed researcher-designed connectives and comprehension tasks. The connectives task was a modified MAZE task, and the comprehension tasks included six constructed response items which spanned literal to inferential questions. (The researcher-designed measures were piloted with a separate sample prior to the intervention.) Covariates included standardized measures of word reading, connectives, and all sub- and composite scores on an English Proficiency test. A language background questionnaire was also administered. A repeated measures ANOVA was used to explore performance and change both within and across instructional conditions.

Results: Findings suggest that students made significant gains in their performance on the connectives task in the third instructional condition, F(2) = 58.29, p < .001. And, across all instructional conditions, contrastive (e.g. but) and causative (e.g. because) connectives posed greater challenges than additive (e.g. and) and temporal (e.g. after) connectives.

Significance: The research on academic language and comprehension is primarily correlational at this point (Uccelli et al., 2015). Causational research linking connectives instruction to comprehension outcomes is scarce at best. The current study offers promising findings on the potential impact of connectives instruction, with the cautionary note that connectives instruction should address the varying levels of challenge inherent in different types of connectives.

Authors

• Dianna R. Townsend, University of Nevada - Reno

• Lizeth Lizarraga

• Monika Bharti, University of Nevada - Reno

• Lisa Johnson, Washoe County School District