Teacher Report and Observation of Classroom Language in Majority DLL Classrooms

• In Event: Better Together: Delving into Multiple Methodologies to Assess DLL Children and Their Classroom Contexts

Abstract

Introduction
Research has demonstrated that support of home language(L1) and English is beneficial for dual language learners (DLLs), across developmental domains including reading, math, and approaches to learning (Burchinal et al., 2012; Chang et al., 2007; White et al., 2020). Despite the importance of L1 support, few studies have examined multiple methodologies to examine classroom language use. Direct observation of language use can be time consuming and invasive. Teacher report offers an alternative to costly and intrusive observation; however, it may inaccurately reflect what actually occurs within the classroom. Studies examining language use in bilingual classrooms demonstrated that teachers were inaccurate reporters of their own language use (Bruenig, 1998; Escamilla, 1992; Irby et al., 2007). Within one study, teachers’ language use also did not align with proportions designated by the district (e.g., 50/50 split reported but 30/70 split through observation) (Irby et al., 2007). These potential discrepancies have yet to be examined within Head Start classrooms which advocate the support of both English and L1 language and literacy skills (HSPS, 2018). Thus, to ensure alignment to policy, there is a need to investigate multiple sources of data to create an accurate picture of classroom language use.
Hypotheses
To address gaps in the literature, the current study examines whether teacher report of language use is related to actual language use in the classroom. These analyses were exploratory.

Methods
A total of 34 lead Head Start teachers (88.20% Hispanic, all female) completed the Center for Early Care and Education Research-DLL Teacher Questionnaire (CECER-DLL; Hammer et al., 2020) in the spring of 2018. The questionnaire includes sections about practices and language use with DLLs.
In the spring, classrooms were filmed across several contexts on a typical day (M=58.18 minutes, SD=12.45). Videos were transcribed and amount of English and Spanish teachers and children used was coded word by word. These counts were aggregated, and total percent was created for the observed amount of English and Spanish that teachers used. Transcripts were also coded for academic Spanish words using a list of words most commonly used/heard by DLLs (Schick et al., 2019). If words were not on these lists, they were deemed “academic”.

Results
Preliminary analyses revealed:
1. Teacher report of Spanish use did not significantly differ from observed use. See Figure 1.
2. Teachers reported that DLLs used more Spanish than was observed.
3. Teachers who reported using L1 vocabulary (never, sometimes, or often; n=10) used significantly fewer academic Spanish academic words than teachers who reported always(n=24) using L1 vocabulary to support DLLs. See Figure 2.
Follow-up analyses with investigate the role of demographic characteristics and teacher training in report of language use.

Implications
Given findings from the current study, teachers could accurately report their own language use, and language support for DLLs. However, it may be necessary to use classroom observation to more accurately depict children’s language use. Implications of drawing upon multiple sources of data to create an accurate picture of classroom language use will be discussed.

Author

• Brooke M Rumper, Yale University
  Presenting Author