Formative Assessment of English Learners' Language Proficiency in the Science Classroom

• In Event: 61.060 - Formative Assessment to Support English Learners' Academic Language Development: Resources for Teachers and Students

Abstract

Perspective & Purpose
The new standards in K-12 education focus less on discrete pieces of knowledge and more on the practices involved in the content areas. The Next Generation Science Standards (NGSS), for example, call for students to engage in language-intensive science and engineering practices such as constructing explanations and engaging in arguments from evidence. Thus, using language is no longer a means to access or demonstrate understanding of science; using language is part of the science construct. The purpose of this paper is to present an approach for formatively assessing English learners’ (ELs) ability to use language in the science classroom in a way that embraces the overlap between the science and language constructs.

Methods
To assess ELs in the science classroom in formative ways, we first defined the construct at the intersection of science and language as the language for communicating science disciplinary meanings. Next, to operationalize this construct, we developed a set of language rubrics to assess students’ responses to science tasks. (See a sample of the language rubrics in Appendix). In operationalizing the language construct, we prioritized meaning. Specifically, we prioritized discourse features, focusing specifically on the language function of the response (e.g., argumentation). Grammatical and lexical features factored into the language score only to the extent that they contributed to or detracted from meaning and precision. Third, to validate these rubrics, we used data from the Science and Integrated Language project, which aims to develop a fifth-grade NGSS-aligned science curriculum with a focus on ELs. We examined responses to science tasks from 100 students—non-ELs, former ELs, and ELs of varying levels of English proficiency—using both a science rubric and our language rubric. Specifically, we examined the distribution of mean language scores across students of varying language classifications and levels of English proficiency. We also compared the relationship between science and language scores based on students’ language classification.

Results
Analyses indicate that language scores based on our rubric are associated with students’ language classification and ELs’ levels of English proficiency. Also, for the most part, responses that were scored high in science were also scored high in language, and those scored low in science were scored low in language. This is consistent with our view that language and science are overlapping constructs. However, in several cases, the science and language scores differed, suggesting that science and language are overlapping but distinct constructs. These discrepancies in particular present opportunities for formative feedback.

Significance
By focusing on the language to communicate disciplinary meanings as opposed to language as traditionally defined (e.g., vocabulary, grammar, organization), content teachers can hone in on those aspects of language that are most critical to engaging in content learning. In the presentation, we will share our language rubrics along with examples of how these rubrics can be applied to students’ responses to science tasks and used by teachers to provide formative feedback that will promote ELs’ science and language learning.

Authors

• Lorena Llosa, New York University

• Scott Grapin, University of Miami