70 likes | 190 Views
Race to the Top Assessment Program General & Technical Discussion. Lizanne DeStefano University of Illinois. Through Course Summative Assessments. Grade level or College/Career Readiness standards/performance levels at the heart of the system, not the assessments;
E N D
Race to the Top Assessment ProgramGeneral & Technical Discussion Lizanne DeStefano University of Illinois
Through Course Summative Assessments • Grade level or College/Career Readiness standards/performance levels at the heart of the system, not the assessments; • Higher education representatives and business community involved in standard setting (content and external validity) • Clear specification of construct/skills to be assessed allows identification of needed access skills for each assessment (construct validity) • Support teaching of target and access skills across grade levels/subject areas • Accessible assessments designed to minimize need for accommodation (e.g. time, print, format, format, reading level, language complexity) (reliability, construct validity); • Special education, ELL and Related Services involved in assessment development
Through Course Assessments (continued) • Multiple assessments should be built with the intent of maintaining high precision across the full performance continuum. • Modular or adaptive assessments aligned to grade level standards and performance descriptors • Predictive and consequential validity of aggregation rules (by subgroups) are central to a sound system.
System for Certifying the Quality and Rigor of End of Course Exams • Once again, rigorous standards that are critical for college or career readiness are central to the validity of any end of course assessment system. • Predictive and concurrent validity evidence is central to determining quality of these assessments.
Computer-based test administration • Novel item types may differentially disadvantage students with disabilities and other subgroups; • Students should have ample experience with technology used in the assessment; • CAT algorithms developed with mainstream population may not adequately assess students with alternate patterns of performance
Innovation and Improvement Beyond 4 Year Grant Period • Huge barrier to effective implementation of NCLB assessment requirement was the lack of time for R&D. • Applicants should submit: • a detailed management plan/timeline including R&D, products, implementation, and use for the four years of funding AND • a second plan describing post-funding activities and benchmarks for full implementation and continued R&D.
Additional, Focused Research • Considerable research exists on growth modeling and value-added methodology. • Studies of use (or nonuse) and consequential validity of various types of high or mid-stakes assessment are needed. • Studies of functioning of the consortia with formative feedback to increase impact of RTT Assessment initiative.