Examining Language Supports for Dual Language Learners (DLLs) in Transitional Kindergarten Classrooms Using the COLES-DLL

• In Event: AERA Roundtable Session Thursday 9:45 am - Gold 2
  In Roundtable Session: Fostering High Quality Early Learning Instruction for Dual Language Learners through Classroom Observation Research

Abstract

Background and Purpose. This study introduces the Classroom Observation of Language and Environmental Supports for Dual Language Learners in Early Learning Settings (COLES-DLL), an observational tool designed to capture language exposure and instructional strategies for DLLs. This work aims to develop and test a tool that is usable across diverse early learning environments, including those with varied language models and student language backgrounds, to inform both research and practice.

Grounded in sociocultural and developmental perspectives, the study emphasizes linguistically and culturally responsive pedagogy and ecological models of teacher–student interactions in classroom contexts (Garcia Coll, et al., 1996; Pianta, et al., 2016). Many widely-used observation tools do not explicitly assess DLL-specific strategies, limiting our understanding of whether DLLs are receiving high-quality, linguistically responsive instruction. Tools like the Classroom Assessment of Supports for Emergent Bilingual Acquisition (CASEBA) have contributed by offering validated ways to observe practices that link to child outcomes (Author, et al., 2019; Freedson, et al., 2014). The COLES-DLL builds on this foundation by offering a complementary approach, emphasizing cycle-based coding of teacher–child interactions across diverse settings and content areas, with particular attention to multilingual classrooms where multiple home languages are present.

Methods. Observations were conducted in 18 Transitional Kindergarten (TK) classrooms (6 dual language immersion [DLI] and 12 non-DLI) using the COLES-DLL. The tool includes: (1) a classroom context survey/interview; (2) cycle-coded observations of teachers’ use of evidence-based strategies for DLLs; (3) ratings of environmental representation; and (4) a checklist of instructional activities and language use. Observers documented the frequency and type of strategies used, the languages, instructional settings and content areas in which they occurred. Data were analyzed descriptively to identify patterns across classrooms.

Results. Findings revealed variation in language strategies, particularly between DLI and non-DLI classrooms. All teachers used a mix of large-group, small-group, and individual instruction, literacy instruction was observed in all classrooms, and math in most. Spanish was the primary home language supported, with limited inclusion of other languages. In Spanish DLI classrooms, Spanish was used consistently across content areas and routines (Figure 1), including for vocabulary instruction in Spanish. In non-DLI classrooms, instruction was primarily in English, though many integrated some Spanish (Figure 1), including vocabulary instruction in both English and Spanish. Teachers frequently modeled language and asked questions, while bridging and other cross-language strategies were less common. Environmental materials in DLI classrooms reflected the target language, while non-DLI classrooms showed limited multilingual representation in the environment (Figure 2), and many home languages were not at all represented in classroom materials.

Significance. This study contributes to research on early childhood education for DLLs by piloting a comprehensive observation tool in TK settings. Based on our initial study using the COLES-DLL, we find that it offers a structured method for capturing nuanced language interactions and environmental supports in multilingual settings. Through future development and testing, we hope to strengthen the tool and ultimately help provide insights for researchers, educators, leadership, and policymakers seeking to improve instructional quality and equity for DLLs.

Authors

• Lisa Jean White, American Institutes for Research

• Leanne Elliott, American Institutes for Research