Performance of Students With Disabilities on New York's Revised Regents Comprehensive Examination in English

摘要

Federal and state policy initiatives are greatly expanding the inclusion of students with disabilities in large-scale assessments, but there is little experience or research to guide this effort. Earlier CRESST studies (Koretz, 1997; Koretz & Hamilton, 1999) examined the experience of Kentucky, one of the first states to include the large majority of students with disabilities in its assessment. The studies revealed a number of important technical and practical issues. State assessments differ markedly, and experience with inclusion may vary from state to state. Accordingly, this study explored the performance of students with disabilities in a field test of the revised New York State Regents Comprehensive Examination in English, the first of the new Regents examinations that almost all students in that state will have to take to obtain a high school diploma. Data from the field test were gathered statewide but not necessarily from a fully representative sample of schools. Accommodations were used liberally, with extra time and testing in a separate location being the most common. Completion rates were similar for students with and without disabilities, and few items had very low p values for students with disabilities. However, students with disabilities scored roughly two thirds to one and one third standard deviations below other students, and a high percentage of students with disabilities provided either unscorable or extremely weak responses to open-response items. The study clearly underscores the need for more extensive information to clarify the effects of including students with disabilities in high-stakes assessments. In addition, it raises concerns about possibly excessive levels of difficulty for some students with disabilities, which could cause either very high failure rates or undesirable responses by teachers or students, such as excessive coaching.