Measuring Academic Language Proficiency in School-Age English Language Proficiency Assessments Under New College and Career Readiness Standards in the United States

Abstract

The current focus across the U.S. on student college and career readiness standards makes clear that both instruction and assessment of academic English will continue to be important for school-age English learner (EL) students. This article presents an overview and summary of key literature on academic language (usually academic English); considers how language demands are represented in the Common Core State Standards adopted by most states; briefly compares how academic English is represented in four sets of current English language proficiency/development standards; and considers some factors and challenges in operationalizing academic English for next-generation English language proficiency assessments. The overview and summary information should be useful to educators and developers of standards, curricula, and assessments who want a deeper understanding of academic language and its significance in EL instruction and assessment.

Notes

¹ See Del Vecchio & Guerrero (1995) for a review of five pre-NCLB K–12 ELP tests.

² Forty-three states and the District of Columbia have adopted the CCSS to date (August 2014).

³ The NGSS are not yet part of the consortia mandate for content-area assessments and are not considered further.

⁴ Naming convention varies by state. This article uses English language proficiency/development (acronym ELP/D) as a way of encompassing the two most commonly used terms, except when referring to a specific set of state standards that uses a particular term.

⁵ The WIDA consortium ELP/D standards were in use in a total of 20 states during the 2008–2009 school year. These standards constituted one set of standards in this analysis. All other states used their own unique ELP/D standards at that time.

⁶ Standards are, arguably, informed to a greater degree by theory than by empirical research. There is something of a distinction among the bases of evidence, then, between those that are more theory-based and those defined by research.

⁷ WIDA is a consortium of 35 states, the District of Columbia, and a U.S. Commonwealth which share a set of ELD standards (WIDA, 2012) and, for a subset of 33, a common ELP assessment (ACCESS for ELLs).

⁸ For a comprehensive review of both theoretical contributions and empirical literature, see Anstrom et al. (2010).

⁹ Findings of this research are often framed in terms of three categories: vocabulary or lexicon, grammar, and discourse features. Language functions, such as describing and explaining, are sometimes considered discourse features and sometimes considered a distinct category. If findings were to be organized in terms of a language proficiency model such as Purpura’s (2004), for example, many of the features listed here as lexical or discourse-level features would be defined in that model as aspects of grammar.

¹⁰ For a thorough discussion of grammatical meaning, see Purpura (2004).

¹¹ This report gives a useful breakdown by grade and content area based on empirical data.

¹² Absent on this list is pragmatic meaning. The empirical research reviewed for this article rarely addressed pragmatics; however, expectations embedded throughout the CCSS that students “adapt their communication in relation to audience, task, purpose, and discipline, … . set and adjust purpose for reading, writing, speaking, listening, and language use as warranted by the task, … . [and understand] how the connotations of words affect meaning” (CCSS Initiative, 2010a, p. 7) suggest that pragmatics knowledge and skills should be more fully integrated into new K–12 ELP/D standards and ELP assessments.

¹³ Another national effort is a framework developed by the Council of Chief State School Officers (CCSSO, 2012) to assist in development of ELP/D standards that correspond to CCRS.

¹⁴ Quinn et al. make this argument in the context of implementation of NGSS.

¹⁵ Arizona, California, Texas, and the WIDA consortium were selected for this review because each has a sizeable number of EL students and each takes a fairly different approach in its ELP/D standards.

¹⁶ The focus in this section is on U.S. state K–12 ELP/D standards. Standards frameworks from other contexts could just as easily be analyzed for language demands. The ACTFL Proficiency Guidelines, the Canadian Language Benchmarks, and the Common European Framework of Reference for Languages, for example, all contain language proficiency descriptors for learners of a second language, and thus provide insight into language demand expectations for language learners in various countries.

¹⁷ Test design for a given assessment should define how measurement of content-area-agnostic and content-area-specific language is to be accomplished, for example, whether a combined approach is to be used, or whether content-area-specific language is to be measured separately for each content area. This design consideration should be informed by the standards to which the test aligns as well as by empirical sources of academic language usage.

¹⁸ As noted, language functions, such as describing and explaining, are sometimes presented in the literature as discourse-level features and sometimes presented as a distinct, fourth category.

¹⁹ The literature has provided little empirical guidance on how grades should be clustered to make up a band. One of the most important issues to be considered in deciding grade bands is the developmental breadth of a given band; some may be wider others. This issue of selecting optimal youngest and oldest grades in a band has implications particularly for standards setting.

²⁰ Comprehension is generally operationalized as a combination of listening and reading comprehension scores.

²¹ CCSS for Mathematics introduce measurement of a circle in grade 7.