Print this Page

Resources

Addressing Two Commonly Unrecognized Sources of Score Instability in Annual State Assessments

Author(s)Gary Phillips
Publication dateFebruary 2011
publication pdfAddressing Two Commonly Unrecognized Sources of Score Instability in Annual State Assessments

Phillips contends that there are two unrecognized sources of error variance in statistics based on state testing data that contribute to score instability: (a) increased error due to sampling of students, or design effects, and (b) error due to equating. Changes in psychometric practice to manage these sources of error variance can increase the ability to detect real change and draw meaningful inferences. The issues and the recommendations for increasing score stability are described in detail.  

Phillips has presented several staff development sessions for TILSA members about these sources of instability and training on statistical software necessary for identifying and controlling for these sources of error, and he has conducted a NCME pre session (2011) on this topic.