students with disabilities compare with other students on many variables related to their schooling and educational achievement. They pointed out the need for research to investigate the relationship between accommodations and validity, not only to document the effects of various accommodations on test scores but also to develop criteria for deciding which accommodations will preserve the validity and usefulness of test results. This committee also noted the need for research to support the development of reliable and valid alternate assessments, and the equating and scaling of such alternatives.
There are even fewer data available for English language learners. For example, simply finding out how many fourth graders across the country are English language learners cannot readily be done with available data. National demographic data on limited English proficient students have been gathered for decades in the U.S. Department of Education’s Survey of the States’ Limited English Proficient Students and Available Education Programs and Services of Education. However, the quality of the data is uncertain because different states have used different criteria for the identification of students of limited English proficiency (summary reports available from National Clearinghouse on English Language Acquisition: http://www.ncela.gwu.edu). An NRC report on the schooling of language minority children identified specific research needs related to the assessment of these children (National Research Council, 1997b, pp. 128-131). It noted in particular the need for research that links assessment strategies to the knowledge base regarding language acquisition, and that can assist in the development of guidelines for when and how to assess these students, and for the development of accurate and consistent means of scoring the work of English language learners.
There is in general, as we have discussed, a lack of research findings that could be used to help policy makers, administrators, and others make decisions about when and how to offer accommodations to students with different kinds of needs. There is a particular need for research into the validity of interpretations made from accommodated scores, and we have described the direction we think this kind of research should take. We would encourage local education officials and schools to undertake such research, to take part in such research when it is conducted by others, and to provide data to researchers engaged in such research. We believe that such research is a crucial element in the effort to build an education system that strives to meet the needs of all students. In order that resources be adequately managed and that the needs of all students be recognized and addressed, those responsible for meeting these goals must have accurate information about all students.