The Language Demands of School is an edited volume describing an extensive empirical base for academic English testing, instruction and professional development. The chapters comprise empirical research by Bailey and colleagues at the National Center for Research on Evaluation, Standards, Student Testing (CRESST) at UCLA, and invited contributions by practitioners in the fields of language policy, testing and instruction. The central focus of the chapters is the research conducted by CRESST over the last two years in an attempt to document the academic English language demands placed on school-age learners of English. The three additional chapters give the perspectives of a policy-maker at the state level, test developers, and practitioners.The Language Demands ofSchool fills a gap in the current literature by addressing the kind(s) of English required of K-12 English Learner students from an evidence-based perspective. This is timely given the broader context of the No Child Left Behind Act of 2001, which has prompted school systems to identify English language proficiency tests to meet the federal mandate. One of the problems that has surfaced in the search for English language tests for K-12 English Learner students is the inadequacy of existing research on the development of the academic English language skills that all students—both English Learner and native English-speaking—need to be successful in the school setting. The Language Demands of School is devoted to exploring this topic and to presenting research that illuminates both the questions and the answers.
|Publisher:||Yale University Press|
|Edition description:||New Edition|
|Product dimensions:||6.12(w) x 9.25(h) x 0.44(d)|
About the Author
Alison Bailey is associate professor and Division Head of Psychological Studies in Education in the Department of Education at the University of California, Los Angeles.
Read an Excerpt
The Language Demands of SchoolPutting Academic English to the Test
Yale University PressCopyright © 2007 Yale University
All right reserved.
Chapter OneIntroduction: Teaching and Assessing Students Learning English in School
ALISON L. BAILEY
The subtitle of this book, "Putting Academic English to the Test," is intentionally ambiguous. Each of the two possible meanings serves an important purpose. In the idiomatic use of the phrase to put to the test, the subtitle invites a challenge to the authors of many of the chapters: Can the research on academic language be put to service in educationally meaningful ways? What does the operationalization of the academic language concept afford us in terms of fairer and better assessment? Will documenting the hallmarks of academic language lead to greater effectiveness in curriculum and teacher professional development? Answering these questions in any great depth would take far more space than can be allotted in this one volume. Rather, the goal of this book is to illustrate what has been achieved so far and to provide frameworks that our research at the National Center for the Study of Evaluation, Standards, and Student Testing (CRESST) has produced so that others might yet yield what still needs to be achieved.
In its second meaning, putting academic English to thetest is read literally-adding the construct "academic English language" (AEL) to assessments of English language development (ELD). This meaning captures the major focus of the book on the current state of language assessment for students learning English in schools in the United States. A review of the available ELD assessments in the spring of 2005 as I write this introduction suggests that much is in flux.
Currently, assessments of English language development aimed at the 5-18-year-old population in the United States fall into three main categories: assessments that have been available for the past 15 to 20 years that appear to tap into the language of social settings, new assessments that have partially integrated the academic language construct into their operational definitions, and assessments under development that have embraced the notion of academic language across all four language modalities-listening, speaking, reading, and writing.
Assessment of English language learners (ELLs) in kindergarten through high school is a requirement of the No Child Left Behind Act (NCLB, 2001). This federal law mandated the yearly assessment of English language skills for ELL students beginning in the 2002-3 school year. Many states are still in the process of responding to this mandate. Many are still administering the older generation of assessments that focus on social or general uses of language rather than language aligned with the discourse of the classroom, textbooks, educational standards, and content-area assessments. Those that are in the process of developing or refining new ELD assessments are daily seeking guidance from language researchers and test developers. The chapters in this book have pulled together much of what my colleagues and I have researched and proposed to a variety of states and the assessment consortia who represent them.
In this initial chapter, I will first provide greater detail about the political and educational contexts that have led to the current activity in language assessment and in assessment and instructional practices more broadly for language minority students in the United States. I will then briefly highlight some of the new assessment development efforts across the states and the implications these efforts will have for instruction and teacher professional development. I hope that the articulation of these implications will make self-evident the need for the detailed review of what we currently think academic English language to be. This review reveals large and educationally critical gaps in our knowledge of what students need to know about language in academic settings and how we think they acquire it.
The Language Mandate
There are major research and policy problems facing the United States and other English-speaking countries with large populations of primary and secondary school students learning English in academic contexts for academic purposes. The problems can be most succinctly articulated as a lack of comprehensive information about what language demands are placed on school-age children in general and how much English language learners can realistically be expected to learn and how quickly. While we have a large literature base about English as a second language (ESL), this research base cannot answer these questions with sufficient specificity to aid policy makers and educators faced with the creation of English language development tests and curricula.
The number of students in U.S. public schools for whom English is a second language has grown steadily during the past two decades. The latest figures show ELL student enrollment at 5.1 million (U.S. Department of Education, 2005). During this period of growth, educators have struggled to implement approaches that help to ensure both quality instruction and valid and reliable assessments of all students. Unfortunately, English language learners often enter school without the requisite English language skills to benefit from the mainstream curriculum, and thus in the past they were often excluded from accountability systems. Now, however, with the No Child Left Behind Act (2001), focus on adequate yearly progress in math, reading, and shortly (2005-6 school year) science for all students (Title I) and special emphasis on ensuring that ELLs make steady progress in acquiring English (Title III) virtually guarantees the inclusion of ELLs in state accountability systems. Under this law, ELL students must show measurable progress each year in the listening, speaking, reading, writing, and comprehension of English (many states are interpreting the fifth element as a combination of the listening comprehension and reading comprehension assessment subsections). Consequently, as already mentioned, states are urgently identifying or developing tests of English language proficiency that can help them meet this federal mandate (Olson, 2002).
Previous research in the area of rate and attainment of ELD has suggested that ELL students can take between four to eight years to achieve the English language proficiency necessary for success on academic content assessments (e.g., Collier, 1995; Cummins, 1981; Hakuta, Butler, and Witt, 2000). Students who start their U.S. schooling in upper elementary grades make more rapid progress than those who start at 12 years and older, often despite the same number of years of exposure to English and years in U.S. schooling (Collier, 1987). Differences in achievement by discipline have also been found, with quicker and greater mathematics attainment (while English language is still developing) than science, social studies (Collier, 1989), or language arts attainment (Butler, Stevens, and Castellon, chapter 2; De Avila, 1997). Relatedly, research on the poorer academic achievement of students who are recent redesignated fluent English proficient (RFEP) students (compared to other students in mainstream classrooms) suggests their difficulties may be due in part to the demands of English in nonsheltered content classes and on standardized content assessments (e.g., Stack, 2002).
While this research base provides estimates of student gains and expectations for ELD performance and its impact on academic achievement, it has limitations for answering questions about the nature of AEL development. These limitations include (1) the appropriateness of existing instruments for measuring forms and uses of school language, as already mentioned above; and (2) the use of primarily cross-sectional design that does not provide longitudinal data on individual student growth and exposure to the language of school. Collectively, these pose a serious challenge to accurate and useful prediction of language and academic outcomes.
Including ELLs in accountability systems is not without challenges. For example, the language demands of content-area assessments may be so great for ELL students as to invalidate the determination of their content knowledge. Ultimately, we often have no way of knowing if the performance of ELL students primarily reflects their language abilities or their content knowledge. Thus, including ELLs in the testing process, knowing that the interpretation of test scores may be invalid, is problematic (August and Hakuta, 1997). However, to exclude ELL students is also unacceptable. If ELL students are not tested, information on their achievement is, in effect, absent from any decision making that impacts their school careers.
Educate or Accommodate?
One widely used means of achieving the goal of inclusion for all students in academic achievement assessment is the use of test accommodations with students who have not yet mastered English (e.g., Butler and Stevens, 1997; Kopriva, 2000; LaCelle-Peterson and Rivera, 1994; Rivera, Stansfield, Scialdone, and Sharkey, 2000). Test accommodations can consist of modifications made to the test itself or modifications made to the test administration procedures, including the test setting. For example, in terms of modifications to the test, students may receive a mathematics test translated into their native language or rendered into simplified English. Accommodations made to the administration of the test include providing students with extra time to complete a test or allowing students the use of English-native language dictionaries.
Studies that have investigated the effects of language accommodations on both ELL and EO students' academic performance (e.g., Abedi, Courtney, and Leon, 2001; Abedi, Lord, Hofstetter, and Baker, 2000; Abedi, Lord, and Plummer, 1997; Castellon-Wellington, 1999; Rivera and Stansfield, 2001) leave us with mixed results as to the fairer assessment of ELL and EO students (extra time, for example, benefits all students). On the one hand, ELL student language abilities, particularly low levels of reading ability, may be a source of construct-irrelevant variance in assessments of academic achievement, such as in mathematics. That is, the mathematics learning that we hope to capture in the assessment may not be reflected; rather, we are measuring the students' inability to read the test questions. On the other hand, if, as Haladyna and Downing (2004) point out, we need to know if students can handle the complexity of language necessary to learn and convey their learning of mathematical concepts, then assessments should include the language demands typical of the instructional contexts students will encounter. Arguably, no less an assessment will allow us to know if students are ready to cope with the language of mainstream mathematics classes.
The complexities of assessing the academic achievement of ELL students are thus many. We argue that there is still much research needed to determine the effectiveness of using accommodations and even then, as others have stated before us, we need to establish clearer procedures for making accommodation decisions in an informed and systematic way (e.g., August and Hakuta, 1997; Bailey and Butler, chapter 4; Haladyna and Downing, 2004). While accommodation studies all share the same objective-to inform educational policy makers about the equitable inclusion of both ELL students and English proficient students on content-area assessments-we have suggested a change of approach that will require future inclusion efforts to begin first by ensuring equitable exposure to and learning of AEL and second by devising academic English language proficiency assessments to help gauge student readiness for mainstream instruction and content-area testing (Bailey and Butler, 2004; Bailey and Butler, chapter 4).
New Assessments of ELD
A major issue, then, at this critical policy decision-making juncture, is how to determine if the requisite English language skills for demonstrating content knowledge on assessments have been acquired by ELL students. Many existing English language proficiency tests do not assess the type of language students must understand and be able to use in the mainstream classroom and on standardized content tests. Existing language tests tend to assess social everyday language rather than the more formal academic English language of the classroom and content tests. Unfortunately for current educational testing needs, basic social language has been found to be only minimally correlated with the more demanding language of school (e.g., Collier and Thomas, 1989).
This is not to minimize the importance of social language skills for successful school and personal outcomes. Indeed, the level of linguistic sophistication necessary to navigate everyday informal situations suggests that we also need to foster student growth in the area of social language development (Bailey, in press). However, currently a student may perform well on a general language test and still not have the necessary language skills for academic tasks (see Butler et al., chapter 2 for a discussion of just such findings with eleventh-grade students). There is, then, an important assessment gap between the type of English an ELL student may know and be able to use-that tested on many current ELD tests-and the language critical to school success.
As mentioned, assessments of ELD fall into three main categories: older assessments that appear to primarily measure social language uses, new assessments that have partially integrated the academic language construct into operational definitions of ELD, and assessments still under development that have adopted the AEL construct as critical to the measurement of scholastic uses of English. For example, the Language Assessment Scales (De Avila and Duncan, 1987, 1988, 1989, 1990; Duncan and De Avila, 1990), are widely used older ELD assessments that are found to have fewer language features matching the language demands of content-area assessments (Butler et al., chapter 2; Stevens, Butler, and Castellon-Wellington, 2000) than derived versions of the LAS such as the current California English Language Development Test (CELDT, CTB/McGrawHill, 2004). The CELDT shows the influence of the California ELD standards (California State Board of Education, 1999) and includes characteristics of classroom language on the test blueprint in all four language modalities (listening, speaking, reading, and writing) (see Bailey and Butler, 2005; Sato, Lagunoff, Worth, Bailey, and Butler, 2005).
Other states are following similar initiatives and have new assessments at different stages of development incorporating varying degrees of the AEL construct. For example, Wisconsin, Delaware, and Arkansas have worked together to produce the WIDA (WIDA Consortium, 2005; see also Davidson, Kim, Lee, Li, and López, chapter 6), which has developed new English language proficiency standards. These will be adopted by Teachers of English to Speakers of Other Languages (TESOL) to replace their original K-12 ELD standards (TESOL, 1997), which had served as the national ELD standards and as the foundation of many state ELD standards and test blueprints.
The English Language Development Assessment (ELDA) is the effort of member states of the Council of Chief State School Officers (CCSSO, 2005). The ELDA will include "topics" from the content areas, such as social studies and science, presumably as part of an AEL construct. Presumably the ELDA item writers will use these topics to create school-related scenarios or to provide specifications for vocabulary to be used in test items rather than to measure content-area knowledge itself.
Both the WIDA and ELDA will include all language modalities. Interestingly, one new ELD assessment responding to the NCLB Act mandate does not include all modalities within its AEL construct. The Stanford English Language Proficiency Test (Harcourt Educational Measurement, 2003), while based on the Bailey and Butler conceptual framework for AEL presented in chapter 4, has articulated speaking skills as social uses of language only and not as part of the AEL construct.
Excerpted from The Language Demands of School Copyright © 2007 by Yale University. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
ContentsForeword by Charlene Rivera....................ix
List of Abbreviations....................xx
ONE Introduction: Teaching and Assessing Students Learning English in School-Alison L. Bailey....................1
TWO ELLs and Standardized Assessments: The Interaction between Language Proficiency and Performance on Standardized Tests-Frances A. Butler, Robin Stevens, and Martha Castellon....................27
THREE Policy Needs: What Federal and State Governments Need from Language Research-Jan Mayer....................50
FOUR A Conceptual Framework of Academic English Language for Broad Application to Education-Alison L. Bailey and Frances A. Butler....................68
FIVE Further Specifying the Language Demands of School-Alison L. Bailey, Frances A. Butler, Robin Stevens, and Carol Lord....................103
SIX Making Choices in Academic English Language Testing: Evidence from the Evolution of Test Specifications-Fred Davidson, Jung Tae Kim, Hyeong-Jong Lee, Jinshu Li, and Alexis A. López....................157
SEVEN Academic English: A View from the Classroom-Margaret Heritage, Norma Silva, and Mary Pierce....................171
EIGHT A Research Agenda for the Future of English Language Development-Alison L. Bailey....................211
List of Contributors....................217