Using Multiple Measures to Enhance Validity and Reliability

Multiple Measures for Assessment and Accountability

September 22-23, 2011 Boston, MA

The call for "multiple measures" in the design of assessment and accountability systems has taken on a mantra‐like refrain. Universally endorsed, federally mandated, scarcely used, and widely interpreted, multiple measures and their appropriate use are at the heart of design decisions for any assessment or accountability system and will be the topic of the 2011 Reidy Interactive Lecture Series (RILS). Teachers have been using multiple measures to award grades forever and measurement researchers have been dealing with this issue ever since the creation of the first composite score. Many technical efforts have focused on maximizing the reliability of the combined score, but there has been much less attention on maximizing the valid use of multiple measures. However, current assessment and accountability initiatives require that we design and implement the most thoughtful approaches possible for using and combining multiple measures of complex student performance to inform instruction and to make overall judgments of “college readiness” or “teacher effectiveness” to name a few current examples.