Assessment of English language learners.

Abstract

The No Child Left Behind Act of 2001 emphasizes the role of high-stakes assessments in reform movements designed to increase accountability for schools and improve achievement for all students. This article describes how English language learners can be included in large-scale standardized assessments, and provides alternative ways to assess them and implications and recommendations for teachers.

Introduction

There are approximately 3.7 million English language learners (ELLs) in U.S. schools (NCES, 2002). They are assessed several times during a school year for a variety of purposes. To find out who qualifies for ESL programs, for placement purposes, to measure progress, to evaluate programs, and to exit ESL. Ideally, tests should yield results that provide information about what students know and can do. But sometimes the same test is used for many purposes or for purposes different than the test was originally intended for. This raises a question of appropriateness and fairness in the assessment procedures for ELLS. Shohamy (2001) claims that tests are powerful tools that are used to make decisions that will impact the lives of the test-takers. Although there are a lot of stakeholder groups involved with a test, only a few of them are affected by the consequences of tests.

Types of Assessments

There are two types of assessments that are commonly used with ELLs: large-scale assessments and alternate classroom assessments. The former refers to standardized assessments that are not developed to assess achievement in a particular program of instruction and are intended to measure more general language skills or knowledge (Genesee & Upshur, 1996). Large-scale tests apply similar standards to all students in a specific grade level. Conversely, alternate classroom assessments refer to assessments that are typically given by individual teachers in their classrooms and are intended to measure a specific skill or knowledge.

Large-scale Standardized Assessments

In the debate on school reform, some favor new methods of assessments to remedy persistent problems in U.S. schools. Many believe that complex, cognitively demanding assessments can drive improvement in curriculum and instruction (Darling-Hammond, 1994). Large-scale assessments are usually mandated for all schools at state or district levels. Large-scale assessments help gather information using standard procedures and this information is used to make decisions regarding the education of ELLs (Gottlieb, 2003). Bilingual education programs routinely require that ELLs take high-stakes standardized tests in English in order to determine their participation in bilingual versus mainstream English classes (Genesee & Upshur, 1996).

Although standardized tests might be helpful to decision makers, they are not necessarily helpful for those directly involved with them (Shohamy, 2001). There is an overrepresentation of ELLs and other students from non-dominant cultures among poor performers on standardized tests (Valdes & Figueroa, 1996). This poor performance is due in part to the lack of full proficiency in English (LaCelle-Peterson & Rivera, 1994). ELLs can achieve minimal oral communicative skills in English fairly quickly, but it takes longer to develop the academic English skills that are needed for academic purposes (Collier, 1987). But the length of time required for full English proficiency is often underestimated. So students may appear to be more proficient in English than they really are and are expected to participate in large-scale assessments before they are fully proficient with the kind of academic language needed to perform well (LaCelle-Peterson & Rivera, 1994). Another problem that might occur when ELLs are assessed using the same tests as native English-speaking students is that teachers might misinterpret their proficiency in reading and in the content areas (Valdes & Figueroa, 1996). Many of these large-scale tests use complex or idiomatic language in the instructions and items, so ELLs might have problems understanding the language of the test or it may take them longer to understand it (Garcia & Pearson, 1994).

Issues of Bias There are huge differences in test scores between mainstream students and minority students (Valdes & Figueroa, 1996). One of the reasons for this disparity is the inequality in educational opportunities and resources (Darling-Hammond, 1994). Additionally, there are significant cultural considerations that are overlooked when assessing ELLs. There is growing evidence that many children do poorly in schools and tests because their cultural frames of reference do not match those of the mainstream cultural reflected in American classrooms (Garcia & Pearson, 1994).

Many of the large-scale assessments that are used to assess ELLs were developed to assess native English-speaking students. Very few ELLs are included in the norming samples of these standardized tests (Davidson, 1994). Since these tests are developed for English-speaking students, the language of the test and of the instructions might be too difficult and challenging for ELLs. They use complex or idiomatic language in the instructions and items. Often, ELLs are unfamiliar with the test language and format (Garcia & Pearson, 1994). Furthermore, many of the tasks, topics and themes included on large-scale tests are not relevant to culturally diverse groups (Garcia & Pearson, 1994). Consequently, many ELLs have little chance to perform well on these types of tests

Inclusion and Accommodation Policies There has been a lot of discussion about the best way to assess the language proficiency and school achievement of ELLs, and how to include them in standards and accountability systems (LaCelle-Peterson & Rivera, 1994). Many large-scale assessments are developed for students who are native speakers of English. Historically, not all ELLs have been included in high-stakes standardized tests (Lara & August, 1996). In order to include all students in large-scale assessments, several accommodation policies are regularly implemented. In principle, these accommodations policies are similar to the accommodation policies found in special education. There are two types of accommodations for ELLs (Butler and Stevens, 2001). Modifying the test (e.g., assessing in the native language, simplifying the language of the test, simplifying test directions, providing graphic support), and modifying the test procedure (e.g., giving more time for students to complete the test, providing oral directions in the native language, using dictionaries). The use of accommodations has been widely proposed as a means of including ELLs in large-scale assessments (Butler & Stevens, 2001).

Accountability Large-scale assessments are closely linked to accountability. It is believed that school achievement will improve if education systems identify what is to be learned and then assess to determine the effectiveness of instruction. The education reform movement in the U.S brought about higher standards along with new assessments in order to help students who were not succeeding. Schools are under a lot of pressure to inform the public about what they are teaching, what students are learning and how well they are performing (Darling-Hammond, 1994). High-stake tests have become the accountability tool of choice in many states as policy makers struggle to find ways to increase student achievement and monitor progress (Gottlieb, 2003). Many schools resort to quick-fix strategies to increase test scores as they feel these competing pressures. LaCelle-Peterson and Rivera (1994) argue that educational reform that serves native English-speaking students does not necessarily benefit ELLs. They argue that tests need to be valid for using with ELLs, normed using ELLs and bias free.

And now with the passage of the No Child Left Behind Act (NCLB), which requires states to implement statewide accountability systems covering all students in the public schools, large-scale assessment becomes synonymous with accountability (Gottlieb, 2003). NCLB requires states to improve the quality of their schools from year to year. Schools and teachers now have the pressure to demonstrate, through large-scale assessments, that ELLs are making improvements in their English proficiency. This is a clear example of how tests are used to determine accountability for student learning based on each state's academic content and achievement standards (Gottlieb, 2003).

Alternative Classroom Assessments

Alternative assessments were proposed as a response to large-scale assessments with the idea that they would enable educators to attend to differences in learners, address learning over a period of time, and include communicative performances in a variety of ways. Alternative assessments are also referred to as authentic assessments or performance-based assessments (Hamayan, 1995). Some of the most commonly used alternative assessments are writing samples, journals, portfolios, classroom projects, and interviews (Hamayan, 1995).

Among some of the advantages of using alternative assessments we find that they are more integrative than traditional tests, are more easily integrated into the classroom, provide easily understood information, are more responsive to each individual learner, promote learning and enhance access and equity in education (Darling-Hammond, 1994). Alternative assessments usually are low-stakes in terms of the consequences and supposedly have beneficial washback effects (Hamayan, 1995). And unlike scores on large-scale assessments, alternative assessments are useful with ELLs because they can provide a multidimensional perspective of student progress and growth overtime (Hamayan, 1995). Among some of the disadvantages of alternative forms of assessments we find that they are not easy to administer and score, are time consuming and lack consistency (Hamayan, 1995). So their use does not guarantee that these assessments are necessarily valid and reliable.

Brown & Hudson (1998) point out that most of the research on alternative assessments are simply descriptive and persuasive in nature and are based on research on empirical studies examining the advantages and disadvantages of the alternative approaches to assessment. They claim that many studies, which advocate for the use of alternative assessments, present their value and validity without providing any evidence to support their claims. Their main point is that these alternative assessments need to also be reliable and valid. They propose a new term, alternative in assessments, to show that there are many different ways to assess ELLs.

Assessment in the Native Language

Assessing ELLs in their native languages is a form of testing accommodation. There are many advantages in using assessments in the students' native languages. For ELLs, future academic success depends on the learners' ability to master English at the same time they are expected to learn content area at their grade level. Yet often, unless assessment of their range of understanding of content knowledge is conducted through their native languages, assessments are highly likely to underestimate their level of past learning (Valdes & Figueroa, 1996). ELLs' low proficiency in English might limit them to express what they already know and what they are capable of achieving. So it is likely that students might be misplaced and put in classes that offer instruction lower than their abilities, becoming disinterested or bored. Similarly, if they are placed in classes that are too high for them, without receiving any kind of support, they will straggle and become frustrated. So using tests in their native languages might provide useful information about their content knowledge.

Additionally, since studies have shown that ELLs transfer literacy skills from one language to the other (Jimenez et al., 1996), literacy assessments in the native language could provide information on the literacy skills the students use and teachers could plan their instruction to help students transfer those skills to English. Unfortunately, most of the assessment instruments in Spanish are simply adaptations or translations of assessment instruments originally developed for native English speakers (Valdes & Figueroa, 1996). This could be problematic because such tests do not have adequate norming, reliability and validity evidence (Garcia & Pearson, 1994). There is a need for developing more tests in other languages, but these tests must be developed for this specific purpose and not simply translated from English versions. These tests need to be field tested and normed with representative samples of ELLs.

Linking Assessment to Curriculum and Instruction

Ideally, we want to develop and use tests that provide information about students. This information would be in turn used to create materials and tasks that are conducive to learning and plan instruction that will meet the needs of all students. Popham (1987) outlined the traditional notion of measurement-driven instruction to illustrate the relationship between assessment and instruction. He claims that assessment directs teachers' attention to content of the tests items, acting as powerful 'curricular magnets'. Proponents of performance assessment believe that what is assessed is what gets taught. If tests affect curriculum and instruction, performance assessment could serve as a driving force for a thinking-oriented curriculum geared toward developing high-order ability and problem-solving skills. New performance-based assessments can be designed to be so closely linked to the goals of instruction as to be almost indistinguishable from them (Popham, 1987).

But Madaus (1988) criticizes measurement-driven instruction mainly because it narrows the curriculum. Teachers only focus on the skills that are on the test and will simply teach to the test. This raises the question about the real purpose of adapting new tests (Shohamy, 2001). Tests are seen as critical agents of educational reform because tests results are relied upon to document the need for change and to demonstrate that change has or has not occurred (Shohamy, 2001). Tests are powerful tools used to change policies and are commonly used as levers of change (Shohamy, 2001). This results in adapting or creating new materials (e.g., books, test preparation materials). Tests also have an effect on all stakeholders to work harder in order to achieve better scores on the test. Davidson and Lynch (2001) describe an approach which involves training teachers in the translation of curricular goals into test specifications. Test specifications specify the intent of a particular test and serve as a link between the standard and the test. They claim that this approach can provide a link between the curriculum, the teacher and the test. This could eventually improve the impact of tests on teaching, especially because NCLB mandates that assessments should be based on state English language proficiency standards.

Implications and Recommendations

Teachers need to develop or select appropriate assessments that reflect the curriculum and the standards used in their classrooms, districts and states. These assessments should allow ELLs the opportunity to demonstrate what they know and can do. But teachers should not base their decisions regarding identification, placement, and progress entirely on a single assessment. Teachers should use multiple assessments to ensure that ELLs are treated fairly and that the decisions that are taken based on these instruments are both valid and reliable, and be completely aware of the consequences that these assessments bear on them and the students. So teachers should be informed of the different assessments that are available for ELLs. Teachers also need to understand that the main purpose behind any assessment is to gain as much information as possible about the students' knowledge and abilities. This includes using the students' native languages in the assessment process.

An issue that is sometimes overlooked in the development of new tests is the issue of really knowing who the test takers are. Test developers should be knowledgeable about ESL and bilingual programs and about the students. Testing makers must gather information about the students' backgrounds, the programs they have access to, how long they have been in these programs, and their proficiency in the first and the second language. This information could help test makers develop better tests. I believe NCLB gives an opportunity to involve teachers in the decision-making process regarding the development of new assessments for ELLs. Test makers need to take advantage of the teachers' knowledge of the English language proficiency and academic content standards, and their knowledge of who these ELLs are.

References

Brown, J. D., & Hudson, T. (1998). The alternatives in language assessment. TESOL Quarterly, 32(4), 653-675.

Butler, F. A., & Stevens, R. (2001). Standardized assessment of the content knowledge of English language learners K-12: Current trends and old dilemmas. Language Testing, 18(4), 409-427.

Collier, V. P. (1987). Age and rate of second language for academic purposes. TESOL Quarterly, 21(4), 617-641.

Darling-Hammond, L. (1994). Performance-based assessment and educational equity. Harvard Educational Review, 64(1), 5-30.

Davidson, F. (1994). Norms appropriacy of achievement tests: Spanish-speaking children and English children's norms. Language Testing, 11(1), 83-95.

Davidson, F., & Lynch, B. K. (2001). Testcraft: A teacher's guide to writing and using language test specifications. New Haven, CT: Yale University Press.

Garcia, G. E., & Pearson, P. D. (1994). Assessment and diversity. Review of Research in Education, 20, 337-391.

Genesee, F. & Upshur, J.A. (1996). Classroom-based evaluation in second language education. Cambridge: Cambridge University Press.

Gottlieb, M. (2003). Large-scale assessment of English language learners: Addressing educational accountability in K-12 settings. Professional Paper #6. Alexandria, VA: TESOL.

Hamayan, E. (1995). Approaches to alternative assessment. Annual Review of Applied Linguistics, 15, 212-226.

Jimenez, R. T., Garcia, G. E., & Pearson, P. D. (1996). The reading strategies of bilingual Latina/o students who are successful English readers: Opportunities and obstacles. Reading Research Quarterly, 31(1), 90-112.

LaCelle-Peterson, M. & Rivera, C. (1994). Is it real for all kids? A framework for equitable assessment policies for English language learners. Harvard Educational Review, 64(1), 55-75.

Lara, J., & August, D. (1996). Systemic refoma and limited English proficient students. Washington, DC: Council of Chief State School Officers.

Madaus, G. F. (1988). The influence of testing on the curriculum. In Tanner, L. N. (Ed.), Critical issues in curriculum: Eighty-seventh yearbook of the National Society for the Study of Education, (pp. 83-121). Chicago: Chicago Press.

NCES (2002). Overview of Public Elementary and Secondary Schools and Districts: School Year 2001-02. National Center for Education Statistics. Retrieved on December 15, 2003 from: http://nces.ed.gov

Popham, W. J. (1987). The merits of measurement-driven instruction. Phi Delta Kappa, 68, 679-682.

Shohamy, E. (2001). The power of test: A critical perspective on the uses of language tests. London: Longman.

Valdes, G., & Figueroa, R. A. (1996). Bilingualism and testing: A special case of bias. Norwood, NJ: Ablex.

Lopez, Ph.D. candidate in Language and Literacy, is a language test developer and an ESL teacher.