Knowing that approximately 25% of children under 5 come from homes where Spanish is the predominant language spoken, we were pleased that Lisa White, a researcher at American Institutes for Research, was willing to speak with us about her study that compared the CLASS with the CASEBA, a tool designed to assess quality in classrooms serving dual language learners. To learn more, read on!
1.) The purpose of your study was to gain a better understanding of how teachers promote learning among Spanish & English speaking DLLs in Head Start. Will you briefly describe your study and tell us why you wanted to examine this topic?
We know that many children in early childhood classrooms come from language backgrounds other than English. These children often have unique needs and strengths when it comes to their learning and development. This means that teachers need to be prepared to implement best practices in instruction for dual language learners (DLLs) when working with them in their classrooms.
To assess best practices in instruction, we typically think of measuring classroom quality. Traditionally, classroom quality has not always measured aspects of instruction that are particularly important for DLLs. For example, are teachers providing slower, simplified talk and using gestures and body language, and a variety of visual supports paired with language when speaking to DLLs? Are they using the home language to support learning? These are best practices as identified in the literature that are critical for DLLs. If we don’t look at quality from a DLL-specific perspective, we will miss out on understanding if teachers are implementing important instructional strategies for their DLLs.
I was particularly interested in examining how teachers were supporting DLLs in these specific DLL practices, but also in general quality instruction, so this study was designed to look at both. We used two observation tools to measure classroom quality: the CLASS to represent general quality, and the CASEBA, Classroom Assessment of Supports for Emergent Bilingual Acquisition (Freedson, Figuera, & Frede, 2009) to represent DLL-specific classroom quality. We looked at how each of these tools related to each other and to child outcomes (i.e., English language skills, Spanish language skills, executive function, and science knowledge) to help understand the aspects of both DLL-specific and general quality that are important for promoting DLL children’s learning.
2.) The CLASS® and the CASEBA both examine classroom quality. Why did you want to compare these two tools?
These two tools were chosen because they are both complementary and unique. Both observations primarily assess aspects of teacher-child interactions, largely in the context of language. The CLASS is so well established in the literature and demonstrates associations with child outcomes, including for DLLs, across Emotional Support, Classroom Organization, and Instructional Support domains, so this was a prime measure to use for general classroom quality. The CASEBA assesses best practices that are specific to DLLs, including how the home language is used in the classroom, so it provides additional and unique information.
In our analyses, we wanted to see if they were related. So, if a teacher provides high quality DLL support, does this also mean they are implementing the critical components of general high quality instruction? Does each tool relate differently to child outcomes? Using the two measures together allowed us to obtain a comprehensive picture of teacher’s language use, specific DLL instructional practices, and general high quality practices, which neither tool captures comprehensively on its own.
3.) In addition to looking at process quality, the CASEBA also examines structural quality (e.g. materials). Can you give an example of some of the process quality items and explain how they support learning for DLLs?
The CASEBA includes many items that are specific to how teachers’ use language and the type of language they use when speaking to their DLL students. For example, teachers are rated on the extent to which they use advanced vocabulary, expand on children’s talk, and have one-on-one interactions with DLL children that are supportive of their language development, depending on their level. These aspects are assessed separately in English and the home language, to measure how these interactions occur in each language. The CASEBA also measures the use of DLL-specific strategies that can help facilitate their English learning, such as how teachers use gestures, objects/visuals, and the home language as a bridge, which are some of the best practices we know are important for DLLs.
4.) You looked at the association between the CLASS and the CASEBA. What did you learn and what are the implications for thinking about quality for DLLs?
The two measures were correlated across domains. In fact, each domain of the CASEBA significantly correlated with each domain of the CLASS. The highest correlation across measures was between the CASEBA Supports for English Acquisition factor and the CLASS Instructional Support domain. Another interesting finding was the correlation between the CASEBA Supports for Home Language domain and the Emotional Support domain of the CLASS, which was among one of the higher correlations across the two measures. These findings may indicate that teachers use the home language of DLLs more when providing emotional and interpersonal support, while English is used more often for academic and instructional talk, which supports some of the existing findings from the research.
In my view, one of the most important findings from this study was the pattern of findings for the Supports for Home Language domain, as they related to child outcomes. Of all aspects of classroom quality measured in this study across the two tools, this domain had the most consistent relationships with child outcomes, predicting Spanish ability, EF development, and science achievement. Although the findings are only correlational, they imply that it is critical to support DLLs in the home language, as it can promote their learning across domains of development.
Taken together, I think these findings show that there are multiple aspects of quality that are important to assess in classrooms with DLLs. Although we see that these different aspects of quality are related across the two tools, they each provide different pieces of information. And, as indicated by the second finding, differential relationships with child outcomes make it clear that using one of these tools is not enough.
5.) Do you have any thoughts on how the tools might complement one another or be used collaboratively to empower teachers and administrators trying to support DLL (and all) students?
I love this question! I think it is so important for teachers to use and explore both tools in practice, to ensure they are providing the highest quality instruction for their DLLs. Most teachers are probably more familiar with the CLASS, given how often it is used for monitoring purposes in many quality rating improvement systems. I think the CLASS can be used as a base to help build their knowledge about what it does and does not measure, when it comes to supporting DLLs. A worthwhile activity could be for teachers to engage in a crosswalk activity to examine how the two tools overlap, and what is unique to both. This could help them see and reflect on what they may already be doing to support DLLs, and where they may need additional support. I would encourage teachers to engage with their teaching teams, and coaches, if possible, and reflect on how they are incorporating high quality practices, from both general and DLL-specific perspectives. In fact, the developers of the CASEBA at National Institute for Early Education Research (NIEER), also have a complementary tool to the CASEBA called the Self-Evaluation of Supports for Emergent Bilingual Acquisition (SESEBA). Teachers and instructional leaders could consider using this tool to explore how to incorporate aspects of DLL-specific classroom quality in their own practice.
6.) What’s next? Where is your research taking you now?
I am currently a researcher at American Institutes for Research. In this role, I am lucky to have the opportunity to continue pursuing DLL work, and am currently working on the F5 California DLL Pilot Study and analyzing a ton of data around instructional practices, professional development, and family engagement for DLLs from a variety of language backgrounds. I also recently helped lead some interesting work on the impact of COVID-19 on DLLs in California; see our two Briefs: California’s Early Learning System and its DLLs During the COVID-19 Pandemic and Supporting Child Care Providers and the DLLs they Serve During the COVID-19 Crisis.
In general, I think future DLL research needs to move towards how to support teachers in multilingual and super diverse settings. The children in my study were all Spanish-speakers, and tended to be in classrooms where Spanish was the only language other than English spoken. But many teachers out there, especially in California, have children from more than three language backgrounds in their classrooms (see here for a brief report on the landscape of early learning programs in California). This language diversity within the classroom has direct implications for how teachers support DLLs of different language backgrounds, and what they can do (or not do) based on if they speak the language of the DLLs they serve. It also has implications for how quality of instructional practice is measured in these super diverse classrooms—teachers may actually provide unique experiences (and therefore, potentially different levels of instructional quality) based on the language background of the DLLs in their classrooms. This is all very complex work, and there is much to do, but we are steadily moving in the right direction, as bilingualism continues to be valued more and more. It is an exciting time to be in this field!
Citation White, L. J, Fernandez, V. A., & Greenfield, D. B. (2020). Assessing classroom quality for Latino dual language learners in Head Start: DLL-Specific and general teacher-child interaction perspectives. Early Education and Development, 31(4). 599-627.