Adaptive Alternate Assessments That Measure Growth

Thursday, June 20, 2013: 10:15 AM-11:45 AM
Chesapeake G-H (Gaylord National Resort and Convention Center)
  • NMAPA Panel 20130325.ppt (2.5 MB)
  • Content Strands:
    1. Transitioning special populations
    The purpose of this symposium is to report the details of a significant improvement in the state-of-the-art in alternate assessment. Five states have adopted an adaptive alternate assessment that is designed to measure growth for students with disabilities (Delaware, Hawaii, Ohio, South Carolina, and New Mexico). The design is also under consideration in Minnesota. These states have formed a collaborative that allows them to share items, reduce cost, facilitate alignment with the Common Core State Standards (CCSS), and seek funding for future research. This Multistate Alternate Assessment Collaborative will represent the third group of states working to improve state alternate assessments. The other two groups of states (or consortiums) are the National Center and State Collaborative (NCSC) and Dynamic Learning Maps (DLM).

    The alternate assessment used by the Multistate Alternate Assessment Collaborative involves the following:

    • The difficulty of the test is adapted to the student’s ability resulting in a more accurate and fair assessment.
    • All teachers and test administrators receive extensive training.
    • Administered and scored by teachers in about one hour.
    • Aligned to the state content standards which in most cases are the Common Core State Standards (CCSS).
    • Meets the technical standards specified in the Joint Technical Standards of the APA, AERA and NCME.
    • Design has been approved by peer review under NCLB (so far in two states-New Mexico and South Carolina).
    • Uses item-response theory models for calibration, scaling, equating and ability estimation.
    • Vertical scale that measures growth across grades (or from fall to spring if administered twice a year).
    • Growth models similar to those used in the state assessment of the general population.
    • Uses the same standard setting methods (e.g., the Bookmark, ID Matching) as used in the state assessment of the general population.
    • Based on universal design principals with many accommodations.
    • Immediate score reports and diagnostic information.
    • Substantially cheaper than many current state alternate assessments.

    The symposium will have one moderator, four speakers and one discussant. Cara Tanimura will serve as the moderator. Gary Phillips will give an overview of the psychometric underpinnings of the adaptive alternate assessment. Tom Dauphinee will discuss and illustrate the measurement of growth for students with disabilities. Brian Touchette will discuss the use of extended standards, performance level descriptors and standard setting with the adaptive alternate assessment. Wendy Stoica will discuss aligning the adaptive alternate assessment to the Common Core State Standards. The discussant will be Rebecca Walk who is an expert in alternate assessments with years of experience from a state perspective as well as a vendor perspective.