A REVIEW OF ISSUES Winter 1998 - Spring 1999 |
The Mid-Atlantic Equity
Center |
By María del Rosario Basterra, Director
National Origin/Language Minority Programs
The following article is based on a testimony presented to the National Assessment Governing Board regarding the inclusion and accommodations for English-language learners on the National Voluntary Tests, by María del Rosario Basterra, Director, National Origin/Language Minority Programs, The Mid-Atlantic Equity Center.
There is little question today about why we need to include English-language learners in all aspects of schooling (including assessment). First, let us consider demographic trends. When teachers look at their classrooms today, they see a much different picture than they did many years ago. Today, one out of three children in the United States is from an ethnic or racial minority group, one out of seven speaks a language other than English at home, and one out of fifteen was born outside the United States (Garcia, 1998). Secondly, educating children of racially, culturally and linguistically diverse backgrounds is a major challenge for school systems across the country. For too many of our language minority students, including English-language learners, American education has not been a successful experience.
Within this context, I would like to focus my testimony today on critical issues regarding the assessment of English-language learners. I will highlight what I consider dilemmas that need to be solved before embarking on the use, and possibly the misuse, of large-scale achievement tests in determining high-stake decisions for these students. What follows is a list of selected issues to be considered as well as a set of recommendations.
Validity Issues
Given the widespread use of largescale achievement tests in making high-state decisions for individual students in states and school districts, Congress wanted to ensure that these tests are used properly and fairly. Therefore, it asked the National Academy of Sciences, through its National Research Council, to Aconduct a study and make written recommendations on appropriate methods, practices and safeguards to ensure that:
existing and new tests that are used to assess student performance are not used in a discriminatory manner or inappropriately for student promotion, tracking or graduation; and
existing and new tests adequately assess reading and mathematics comprehension in the form most likely to yield accurate information regarding student achievement of reading and mathematics skills.
How accurately can standardized tests in English assess English-language learners knowledge? How valid are the results of these tests? Should we focus only and primarily on standardized tests when assessing these students?
Let us first discuss the questions related to validity. As noted in the Standards for Educational and Psychological Testing, every assessment is an assessment of language (American Educational Research Association, American Psychological Association, and National Council of Measurement on Education, 1985). It follows that if a student is not proficient in the language of the test he/she is presented with, his/her performance on the test will be affected by construct-irrelevant varianceC that is, his/her test scores will likely underestimate his/her knowledge of the subjects being tested.
In addition, and as indicated by the National Research Council (1997), there are additional problems related to the validity of many of the most commonly used standardized tests. These include: 1) a norming bias (small numbers of English-language learners in the sample, making it potentially unrepresentative); 2) a content bias (the test reflects the dominant-culture standards of language, knowledge and behavior); and 3) a linguistic and cultural bias affecting students' formal test performance (timed testing, difficulty with English vocabulary, and the great difficulty of determining what bilingual students know in their two languages).
Even though many of the validity concerns could and should be addressed in order to make non-referenced tests more accurate, there is no doubt that unless the students taking the test are proficient in the language of the test, their performance will only represent a modest sample of their true knowledge. Even when standardized tests are translated, they will still not represent a holistic picture of what these students really know.
Taking these issues into consideration, should we focus only, and primarily, on standardized tests when assessing these students?
Criterion-reference tasks in general, and rubrics in particular, can provide a more comprehensive and accurate picture of English-language learners knowledge than one single snap shot score provided by a norm-reference test. While the language proficiency factor will also affect students= scores on alternative forms of assessment, such as performance assessments, we believe that they provide English-language learners with opportunities to demonstrate their knowledge in a variety of ways, including non-language based tasks. In no way are we trying to imply that standardized tests are not useful. If properly constructed they can provide important information about what English-language learners know and how they are performing in comparison to other groups.
However, it is important to remember that even the best constructed norm-referenced tests (even when properly translated) will only provide limited information about what these students know.
The use of testing results: Testing for learning or testing for sorting?
Another critical issue in relation to standardized tests is the possible misuse of them. We need to remind ourselves that despite warnings by researchers and educators against the use of single indicators of performance in academic achievement, most high- stakes decisions in education are still being made on the basis of a single test score typically obtained from standardized norm-referenced test.
While the information provided by standardized tests can and should be used to inform teachers, students and parents on the needs and strengths of student performance, they are most often used to sort, track and determine promotion and graduation. I believe that while this practice is not valid for any student, it is particularly discriminatory for English-language learners for the reasons mentioned above.
Assessment and accountability: Are the results of standardized tests the only way to hold schools accountable for English-language learners' performance?
While it is true that the previous practice of not including English-language learners in state and other large scale assessments resulted in not making schools accountable for these students, the current practices of making accommodations to include larger number of English-language learners does not necessarily solve the problem.
First, large scale tests are seldom used to improve instruction. They are mostly used to sort and track students.
Second, only a imited number of English-language learners are included in the large acale assessments.
Third, the results of these tests are not only limited but in many cases non-valid since they do not properly reflect students' knowledge.
Fourth, in many states the data is not desgregated by English-language proficiency level; therefore, it becomes almost impossible to make schools accountable for the specific performance of these students.
What is needed is a more comprehensive system that ensures that English-language learners are evaluated with appropriate and valid assessments that include both norm-referenced and performance-based assessments. Specifically, if states develop criteria for high-stake decisions aligned with their content and performance standards, they should develop ways to validate and align existing instruments or to develop parallel valid measures and scoring rubrics for English-language learners. If large scale standardized tests are to be used, efforts should be made to improve the validity of tests; to assess students in their primary language; and to make appropriate accommodations in the administration of the tests.
The accountability system should take into consideration the results of multiple assessments. Decisions regarding the placement, promotion and graduation of English-language learners should never be the result of one single test. They should be based on a series of valid assessments that take into consideration their level of proficiency in English as well as their culture and language background.
Regardless of the complexity of the assessment process, and regardless of how difficult it might be to assess English-language learners, school districts must develop systems of accountability that fully incorporate linguistically and culturally diverse students. Even if it is not appropriate for students to be assessed in English, and even if no native language assessment procedure are available, schools should have alternative ways of ensuring that these students are being given the best opportunities to achieve the standards set forth for the rest of the student body. It is clear that properly assessing English-language learners is a complex and challenging task. However, it needs to be done. Otherwise, we will not only be in violation of the most basic evaluation principles, but also openly discriminating against this group of students.
For more information regarding our services, please contact Sheryl Denbo at 301-657-7741 ext. 102 or visit our website at www.maec.org
References
American Educational Research Association, and National Council on Measurement in Education. (1985). Standards of Educational and Psychological Testing. Washington, DC: American Psychological Association.
García, E. (1998). Excellence and Equity for Language Minority Students: Critical Issues and Promising Practices. The Mid- Atlantic Equity Center, Chevy Chase, MD.
National Research Council (1997). Improving Schooling for Language Minority Children. D. August and K. Hakuta, editors. Board of Children, Youth, and Families. Washington, DC: National Academy Press.