Correcting Fallacies in Validity, Reliability, and Classification

This article reviews three topics from test theory that continue to raise discussion and controversy and capture test theorists' and constructors' interest. The first topic concerns the discussion of the methodology of investigating and establishing construct validity; the second topic concerns reliability and its misuse, alternative definitions of reliability, and methods for estimating reliability; and the third topic concerns the relationships between reliability, test length, and the insufficient quality of decision making using short but reliable tests.