Abstract
Test items undergo multiple iterations of review before states and vendors deem them acceptable to be placed in a live statewide assessment. This article reviews three approaches that can add validity evidence to states' item review processes. The first process is a structured sensitivity review process that focuses on universal design considerations for items. The second method is a series of statistical analyses intended to increase the limited amount of information that can be derived from analyses on low-incidence populations (such as students who are blind, deaf, or have cognitive disabilities). Finally, think aloud methods are described as a method for understanding why particular items might be problematic for students.
Original language | English (US) |
---|---|
Pages (from-to) | 25-36 |
Number of pages | 12 |
Journal | Educational Measurement: Issues and Practice |
Volume | 27 |
Issue number | 1 |
DOIs | |
State | Published - Mar 2008 |
Keywords
- Assessment
- Disability
- Large-scale assessment