Learning Connectives in Second Grade: Student and Word-Level Factors That Contribute to English Learners’ Progress

• In Event: Roundtable Session 24
  In Roundtable Session: The Development of Vocabulary and Comprehension in Spanish-Speaking English Learners

Abstract

Objectives. This study presents promising results for an innovative intervention designed to teach connectives to second-grade Spanish/English bilingual students with emerging English proficiency. Student factors (vocabulary knowledge) and word-level factors (word difficulty) were explored as mediators of the intervention impact.

Theoretical Framework. Research has highlighted the need of explicit vocabulary instruction for children with low English proficiency from bilingual families with low economic and educational resources (August & Shanahan, 2008). Multiple vocabulary interventions have been designed as a response to this well-documented pedagogical need, yet most efforts have targeted content words (nouns, verbs). Non-content words, such as connectives, have been usually overlooked. Connectives (e.g., because, however) are function words with minimal lexical meaning that mark inter-sentential relations. Research indicates that connective knowledge is positively associated with reading comprehension in L1 and L2 for children and adults (Pretorius, 2006). Knowledge of connectives has been identified as an area of particular vulnerability for 4th-grade Spanish-speaking English learners in U.S. classrooms (Crosson, Lesaux, & Martiniello, 2008). Guided by research-based principles of vocabulary instruction (Stahl & Nagy, 2006), the Connect intervention attempts to fill a gap in the menu of currently available vocabulary programs by exploring a multimodal approach to teach connectives to young learners.

Methods. Using high-quality, culturally-relevant and age-appropriate children’s literature as the context for instruction, activities were designed to provide multiple means of expression and representation, and numerous opportunities for language production in playful communicative activities in which teachers gradually release their instructional support. Twelve connectives --prevalent in primary-school texts yet not known by participants-- were selected for instruction (Uccelli, Rosenthal, & Barr, 2011). Connectives were stratified by conceptual relation and word difficulty (see Appendix). In the treatment group, teachers taught connectives explicitly during an eight-week period. The control group was exposed to the same books and connectives but without explicit instruction. Bilingual Spanish/English second graders from high-poverty schools participated in the study. The treatment group included 228 students (12 classrooms); the control group had 196 students (ten classrooms).

Data Sources. Sources include the Test of Oral Language Development (TOLD): vocabulary subtest and the Connective Assessment: researcher-designed test (previously pilot tested for validity and reliability).

Results. To investigate differential gains by word difficulty level and initial levels of students' vocabulary, we conducted a repeated measures ANOVA with time (pre/post Connective Assessment) and word difficulty (level 4 or 6) repeated within student being predicted by instructional condition (treatment vs. control) and students’ initial vocabulary. Results showed a significant 4-way interaction for time, condition, word difficulty, and vocabulary. For level-4 words, students made about an 18% gain regardless of treatment or vocabulary level. However, for level-6 words, treatment students with high-vocabulary made significantly greater gains than low-vocabulary treatment students (33% vs. 11%), as well as high-vocabulary control students (6%) and low-vocabulary control students (11%).

Significance. Results highlight the need for more comprehensive vocabulary interventions, revealing the malleability of connective knowledge in young children.

Authors

• Christopher Daniel Barr, University of Houston

• Paola Uccelli, Harvard University

• Franne Rosenthal, Harvard University