| Sign In to gain access to subscriptions and/or personal tools. |
DOI: 10.3102/0013189X07311600 © 2007 American Educational Research Association Construct Validity: A Universal Validity System or Just Another Test Evaluation Procedure?Professor of psychology and director of quantitative psychology at the Georgia Institute of Technology, School of Psychology, 654 Cherry Street, Atlanta, GA 30332; susan.embretson{at}psych.gatech.edu. She was the recipient of the 2001 Distinguished Scientific Lifetime Contribution Award, American Psychological Association, Division of Measurement, Evaluation and Statistics. Her research interests include measurement, psychometric methods, cognition, and individual differences Lissitz and Samuelsen (2007) have proposed a framework that seemingly deems construct validity evidence irrelevant to supporting educational test meaning. The author of this article agrees with Lissitz and Samuelsen that internal evidence establishes test meaning, but she argues that construct validity need not be removed from the validity sphere. In fact, she argues that doing so could have an adverse impact on the quality of educational tests. She proposes a universal system for construct validity to illustrate how diverse evidence is relevant to claims about measuring examinees knowledge, skills, abilities, and competencies even when test specifications provide a major source of evidence.
Key Words: construct validity content validity educational testing
This article has been cited by other articles:
|
||||||||||||||||||||||||






