Haladyna, T. M., & Downing, S. M. (2004, March). Construct-irrelevant variance in high-stakes testing. Educational Measurement: Issues and Practice 23(1). Retrieved January 3, 2007 from http://www.blackwell-synergy.com/doi/pdf/10.1111/j.1745-3992.2004.tb00149.x
The authors define and examine the common threat of construct-irrelevant variance (CIV) to validity, the appropriate interpretation of test scores, and its sources and pervasiveness in high-stakes testing. The authors argue that education leaders need to pay increased attention to this serious problem in order to improve the effectiveness and validity of evidence for high-stakes testing programs.
Analysis, Data Quality Assurance, Quality of Assessment Data
The DataUse web site is a part of AACC
and was created and is regularly updated by CRESST
in partnership with WestEd,
and supported by the U.S. Dept. of Education.
